動議 · 2026-05-06 · 屆國會 15
人工智慧（AI）轉型不容無就業式增長（主辯論）
5月6日國會續辯職總秘書長黃志明提出的"人工智慧轉型不容無就業式增長"動議，約20名議員發言，是第15屆國會迄今最具分量的AI辯論。動議要求承認AI的經濟變革力量，把增長錨定在公平、韌性與人人有機會，裝備工人與企業，確認新加坡不容無就業增長。執政黨與工運議員聚焦崗位重設計、公司培訓委員會（CTC）與新設的三方就業理事會；工人黨議員全數支援動議但另提結構方案：嚴燕松倡全民AI股權基金（成年公民每年500新元分紅加在職培訓基金），Andre Low主張無收入上限的裁員保險、再培訓稅收抵免與"AI收益審計"，Kenneth Tiong呼籲高階AI工具全民可及。人力部長陳詩龍駁斥工人黨方案是"和解金"而非賦權，引人力部調查稱採用AI企業僅約6%裁員，承諾研究提高求職者援助收入門檻並推動更早裁員通報。朝野最終均表態支援動議。
關鍵要點
- • 約20名議員發言近7小時，朝野全數表態支援動議，但在政策工具上針鋒相對
- • 工人黨提全民AI股權基金（成人每年500新元分紅）與無收入上限裁員保險
- • 陳詩龍引人力部調查：採用AI企業僅約6%裁員，七成已見生產力提升
- • 政府允研究提高求職者援助收入門檻，推動僱主在員工離職前通報裁員
政府拒絕以現金再分配為主的保障路線，堅持"投資於人"：通過技能與勞動力發展局、三方就業理事會和企業勞動力轉型配套讓工人隨AI增長共同進步，承諾"保不住每一份工作，但會保護每一個工人"。
工人黨議員支援動議本身，但批評現行求職者援助計劃門檻過低、遞減設計逼工人倉促就業，主張無收入上限的裁員保險、全民AI股權基金、再培訓稅收抵免與強制性AI轉型通知等結構性保障。
AI就業衝擊已升為國家級議程：政府明確以三方機制而非立法保障應對AI轉型，公共資助將與工人成果掛鉤，求職者援助計劃門檻與裁員通報制度料將檢討增強。
“你們兩位的提案都不是賦權。在我看來，那是一種和解金——認定大規模崗位流失不可避免，我們能做的最多隻是減輕衝擊。”
參與人員 (12)
完整譯文（中文）
Hansard 原始記錄 · 2026-06-09
（程式文本）恢復就問題進行的辯論[2026年5月5日]，（程式文本）
（程式文本）本院——（程式文本）
[(議事文本)1. 承認新技術特別是人工智慧(AI)的變革力量推動新加坡經濟發展的下一階段；(議事文本)]
（程式文本）2. 強調新加坡的人工智慧驅動增長方案必須以公平、韌性和全民機會為基礎；（程式文本）
[(議事文本)3. 決心為工人和企業提供裝備和支援，以抓住新機遇並共同前進；以及(議事文本)]
（程式文本）4. 確認經濟進展必須保持包容性，新加坡不能出現無就業增長，因為每一名工人都很重要。（程式文本）
（程式文本）問題再次提出。（程式文本）
下午12時32分 議長：Mark Lee 先生。
Mark Lee 先生（委任議員）：議長先生，基於我的同事、全國職工總會（NTUC）秘書長Ng昨晚所做的演講，我想著重關注這項議案如何在企業層面真正實現。
先生，防止無就業增長不能僅依靠工人支援來實現。它還必須融入企業轉型的方式、崗位重新設計的方式以及我們的體系支援企業自信前行的方式。
歷史表明，在技術變革時期，贏家不是那些試圖保護現有商業模式的企業，而是那些充分理解自身基礎優勢並能夠將其重新部署到新領域的企業。Fujifilm就是一個例子。它沒有通過試圖銷售更多膠片來渡過膠片行業的崩潰，而是基於材料、光學和成像技術的能力，進入了護膚、診斷和醫療保健技術領域。
在新加坡已具有優勢的行業——先進製造業、物流和連通性、金融和醫療保健——人工智慧並非在替代產業本身，而是改變了價值在產業內創造的方式。這是新加坡現在需要的思維方式。我們的任務不是逐一保留現有的每一份工作。我們的任務是幫助我們的企業和勞動力識別他們的優勢，適應這些變化，並攜手進入增長的下一階段。
我的發言將分為三個部分：首先，人工智慧如何實際上為企業創造價值；其次，它對就業和工作者的影響；第三，我們如何能構建一個更清晰的企業入口和一座更寬的橋樑，使企業和工作者能夠共同渡過轉變。
議長先生，企業採納人工智慧是因為它能帶來可衡量的成果。證據已經顯現。在客戶服務領域，生成式人工智慧將生產力提高了約15%，在經驗較少的工作者中獲益更大。企業在軟體工程方面也看到了兩位數的增長，而在運營和供應鏈中，人工智慧正在改善預測、減少浪費並最佳化庫存和物流。
其次，人工智慧不僅僅改進工作。它重新配置工作。全球證據表明，人工智慧的採納越來越集中在高技能工作者中，企業在由人工智慧支援的較小、更有經驗的團隊周圍重新組織工作。這造成了雙重風險：首先，生產力收益可能更多地流向已經處於領先地位的人，加劇不平等；其次，它有可能侵蝕職業生涯的底層。
如果入門級職位減少過快，工作者通過經驗積累判斷力的傳統途徑將被打破。企業可能仍然需要中層能力，但較少有工作者會有機會培養這種能力。這是一個結構性問題。
與此同時，新角色正在湧現，包括將人工智慧整合到工作流程中、驗證輸出、重新設計工作和將領域知識轉化為解決方案。DBS和Mastercard等公司正在使用人工智慧來處理日常查詢並大規模個性化響應，將人工代理釋放出來從事更高價值的工作。我們在中小企業（SME）中也看到同樣的情況。例如，護膚公司MTM Labo使用名為Hana的人工智慧工具來支援多語言客戶諮詢，使其團隊能夠專注於更復雜的、高接觸度的互動。
這是一個重要觀點。人工智慧在思慮周全的部署下，不僅僅是替代工作。它改變了工作的性質，並能提高人類工作的價值。
但採納並非即插即用。它需要整合到工作流程中、流程重新設計以及與商業戰略的協調。這是許多企業，特別是中小企業，在將人工智慧轉化為實施時面臨的挑戰所在。如果我們不解決這個問題，能力將集中在較大的企業和更高技能的工作者中。差距將擴大。這一結果將直接違背這項議案的精神。
然而，我們也必須謹慎對待如何應對這些變化。
我理解Ng先生要求提早通知裁員以更好地支援工作者的良好初衷。但如果企業還沒有準備好重新設計工作或以不同的方式安置工作者，僅提早通知是無法解決問題的。
如果人工智慧正在重新組織工作，也許更加重要的更好解決方案不是我們是否在失業後更早介入，而是在失業成為必要之前足夠早地介入。我們應該將重點從管理裁員轉向使所有企業都能夠重新設計工作和對工人進行再培訓，使得他們的轉變和勞動力調整同時進行，而不是在轉變之後。
這引出了我的第三點——我們如何架起前路的橋樑？一座足夠寬闊使許多人都能通過的橋樑，而非僅供少數人通過的橋樑。
最近推出的企業勞動力轉型計劃（Enterprise Workforce Transformation Package）是正確方向上的一步。在新加坡工商聯合會（SBF）和新加坡全國僱主協會（SNEF）作為專案合作伙伴的情況下，企業可以獲得諮詢支援來重新設計流程和工作角色，並利用SkillsFuture勞動力發展補助金（SkillsFuture Workforce Development Grant）進行諮詢、勞動力技術採納和能力建設。這是一個有意義的轉變。
然而，目前，企業仍在應對多項計劃、多個機構和多個審批流程，在速度最為重要的時刻放緩了採納。但我很欣慰聽到Tan部長昨晚所說，SkillsFuture新加坡和勞動力新加坡將合併為技能和勞動力發展局（Skills and Workforce Development Agency），旨在解決這個問題。
但我們能否為幫助企業做更多的事呢？
這涉及我討論的實用性的第一個方面。前進的一種方式是，人工智慧補助和計劃採用更加綜合的方式，使企業可以通過單一介面獲得這種支援，而不是應對多個機構，並靈活地部署它，無論是用於基本實施、訂閱還是實驗，而無需重複的審批層次，就像SkillsFuture企業信用錢包（SkillsFuture Enterprise Credit wallet）一樣。
對於規模更大或更復雜的定製專案，也可能有理由提供更多的前期支援，以便企業在進行長期投資時不受現金流的制約。
與此同時，我們必須認識到人工智慧的採納涉及實驗。並非每個專案都會成功。如果企業儘管真誠努力仍然受到懲罰，我們有可能會阻礙創新。如果我們希望企業果斷行動，對失敗的合理容忍將是必要的。
第二個維度是規模化能力建設。經驗豐富的人工智慧專業人士的人才庫仍然很小且競爭激烈。如果我們對引進外國人工智慧人才過於限制，我們將減緩整個經濟的能力建設。因此，我們為打造寬廣橋樑所做的努力也必須意味著保持我們的人才管道開放。不僅是為了引進人才，而且要讓這些人才在各企業和行業間進行技能交叉流動。
對於中小企業，我們可以考慮的一個領域是提供有針對性的靈活性，以引入超越現有人力資源限制的專門AI專業知識。這可以是有時間限制的、按申請基礎進行的，並配備一些措施防止濫用。
第三個維度是通過我們的高等教育學院（IHLs）。高等教育學院可以更刻意地定位為應用AI的執行平臺，特別是通過以研究生課程為中心的卓越中心。許多研究生，包括外國人才，帶有豐富的行業經驗和技術深度。與本地本科生配對時，形成了一個實用的能力轉移模式。如果以真實的中小企業和部門問題為中心，這些團隊可以超越概念驗證工作，開發可部署的解決方案。
這實現了幾個目標。它降低了中小企業的實驗成本，同時促進了國際和本地人才之間的交叉融合，並建立了一個初創企業的管道，以新加坡為中心，專注於解決真實的行業需求。
這個模式也有助於應對職業發展階梯斷層的問題。隨著入門級途徑變窄，讓學生參與真實問題解決可以更早地建立能力。與此同時，它強化了我們學生的批判性思維，使他們準備好質疑和驗證AI，而不是僅僅依賴AI。
第四個維度是行業賦能。今天，許多公司在孤立地試圖解決類似的AI問題。這導致了努力的重複、更高的實驗成本和更緩慢的採用。
行業協會和商會在結構上處於解決這個差距的位置。它們在政府政策和企業級行為的介面處運作，可以將國家AI戰略轉化為特定部門的實施。它們可以充當協調平臺，識別共同的行業問題，匯聚需求，並與解決方案供應商和高等教育學院合作，開發與實際工作流程和職位相符的綜合、可部署的解決方案。
讓我舉例說明。新加坡商人聯合會正在開發一個AI工具，以幫助企業理解自由貿易協定的原產地規則，以及在商品出口到海外時如何申請優惠關稅待遇。這是一個行業主導方法如何減少重複和提高效率的例子。
對於許多中小企業，挑戰也在於應用能力。企業想知道應該優先考慮哪些問題，誰可以幫助以及如何在不過度成本或風險的情況下進行。
新加坡中華總商會（SCCCI）與信息通信媒體發展局（IMDA）聯合推出的AI體驗計劃作為支援數字企業藍圖的舉措，清楚地說明了這一點。它被過度申請是因為中小企業正在尋求有指導的入口點。
同樣，SCCCI的AI賦能計劃允許中小企業定義真實問題，並與南洋理工學院、新加坡理工學院、淡馬錫理工學院和新加坡科技設計大學（SUTD）的學生合作開發解決方案。中小企業獲得了可行的解決方案。學生獲得了相關的經驗。知識在整個系統中傳播。有了適當的支援和資金，行業協會和商會可以成為加速各部門AI採用的平臺。
最後，議長先生，第五個也是最重要的維度是人力資本。AI驅動增長的成功將不僅由我們部署多少工具決定，還由我們引導多少工人度過這種變化決定。
我同意吳先生的觀點，無論是由於AI還是行業整合而被替換的工人在過渡期間都需要更強的支援。但我們也應該考慮該支援如何構建。
今天，許多都遵循'先培訓後安置'方法，其中工人首先被重新培訓，然後被支援找工作。實際上，這可能是不確定的。沒有工人想被裁員，花幾個月時間培訓，然後仍然面臨下一份工作的不確定性。因此，我們應該更刻意地朝著'先安置後培訓'模式前進。如果另一家公司準備接納一名被替換的工人，即使適應不是立即的，我們應該直接支援這種過渡。這可以通過向接納企業提供臨時工資支援來完成，類似於工作支援計劃（JSS）的精神，但以行業過渡基金為中心，使公司有動力優先招聘然後進行在崗培訓。
這縮短了工人的不確定性期間，同時給公司信心接納和培養新人才。這也是行業協會和商會以及工會可以發揮協調作用的地方——識別有需求的公司，並將其與面臨風險的工人相匹配。
最後，議長先生，開放的橋樑必須是道德和社會契約，該契約的核心是信任。工人必須將AI視為賦能工具，而不是威脅。如果AI被視為消除工作或關閉途徑的工具，採用將減緩，不是因為公司缺乏技術，而是因為缺乏信任。
如果信任被打破，我們可能會無意中創造一個雙輸的結果，即政府幹預更受限制的關於AI採用的勞動監管，提高長期成本並減少企業的靈活性。因此，必須通過AI的部署方式、工作的重新設計以及工人如何度過變化來刻意建立信任。
先生，我是一個商人，我加入支援我的工會兄弟吳先生在這項動議中的號召，因為我相信這體現了為新加坡服務良好的真正三方合作精神。
我們不能說每個工人都重要，然後讓工人獨自度過這種過渡。企業必須領導重新設計工作和投資於其人民。工人必須向前邁進並適應。政府必須確保系統使兩者都能夠。只有這樣，工人和企業才能共同進步。
議長先生，這項動議反映了新加坡決心正確處理這個問題。讓我們採取積極的道路共同合作，儘早建立信任，並確保AI擴充套件機會而不是縮小機會。先生，我強烈支援這項動議。[掌聲。]
議長：Saktiandi Supaat先生。
下午12點49分 Saktiandi Supaat先生（Bishan-Toa Payoh）：議長先生，在我開始之前，我想宣告我在新加坡的一家銀行、金融機構工作。我還是電力和天然氣僱員工會（UPAGE）和物流和供應鏈工會（SCEU）的顧問。
議長先生，我首先要感謝全國工人大會秘書長兼議員吳慶銘先生提出這項重要動議，並闡述了需要一個新的AI驅動增長契約，該契約將工人置於我們轉變的中心，以公平、韌性和全民機會為基礎。
我將專注於我們如何建立一個更具包容性的AI經濟，其中增長不僅強勁，而且廣泛共享。
AI不再是新興的；它已經在重塑我們的工作和生活方式。除了眾所周知的AI工具，我在與UPAGE和SCEU的合作中，以及最近人力資源政府議會委員會對SMRT、NTUC Finest at Punggol和Chye Thiam Maintenance Pte Ltd的學習之旅中親眼見證了這一點。
在不同的部門，電力和天然氣、供應鏈、運輸、零售、食品和飲料（F&B）、清潔和設施管理，所有這些都在工作流程中嵌入AI、自動駕駛車輛（AVs）和機器人，提高生產力，建立新工作，為現有工人重新設計角色。
現在存在AI創新和應用的全球競爭。並且越來越相信那些早期和果斷行動的經濟體和公司將捕獲最大價值。
但議長先生，正如本議院一直強調的那樣，經濟成功必須是包容性的。經濟成功不僅僅是增長，而是其利益如何廣泛分享。不均勻的增長不是新加坡應該遵循的模式。除了我們通過稅收措施和有針對性的援助實施的再分配措施外，我們必須培養一種持久的心態，以確保AI驅動的增長是包容性的。
AI的使用引發了許多工人的關注。例如，AI會奪走我的工作嗎？它會導致無工作增長嗎？
根據全國工人大會每年進行的新加坡經濟情感調查，AI會替代他們的工作或當前職位的恐懼對專業人士、經理和技術人員（PMETs）和入門級求職者更為明顯。對於非PME和低薪員工，關注程度較低可能是因為他們不廣泛使用AI工具，並且不瞭解AI將如何影響他們的工作機會，而不是因為不存在AI破壞風險。
這些關注得到全球工作重組的新聞以及AI在不同部門和職業間採用不均衡的加強。
尊敬的議長先生，讓我用我所在的金融部門的具體例子來說明這種不均衡。在新加坡的銀行業，人工智慧不再是新興技術，而是已經深度融入併成為生產力的核心驅動力。在整個部門，我們本地的銀行正在運營和客戶服務中部署自動化，特別是在勞動密集型和重複性流程中。這些不是試驗性用例；它們正在改變整個價值鏈上的工作方式。
人工智慧現在被用於運營中，如處理、結算和合規，以及通過支援更快速和更一致承保的機器學習模型進行信用評估。它甚至開始——好吧，我不會說是開始；它已經在招聘中發揮更大的作用，特別是在初始篩選中。
但這種影響並不是均勻的。日常文秘和處理工作受影響最大，而招聘正在轉向資料、人工智慧、網路安全和治理方面的更高技能崗位。與此同時，監管正在維持對風險、合規和審計監督角色的需求。
但最重要的是，更復雜的工作仍然需要人類的判斷。處理細微的客戶問題、管理關係和做出判斷決策不能輕易自動化。因此，許多職位不是在消失而是在演變。例如，信用官員正在轉向解釋和監督，而客服中心的角色正在轉向體驗、升級和建立信任。所以，我們看到的不是大規模的工作替代，而是工作中任務的重新配置。
尊敬的議長先生，雖然許多人工智慧討論關注PMETs，但我們不能忽視熟練的技工和我們的藍領工人。我們的電工、技術人員和維護工人是必不可少的。人工智慧無法獨自修理電梯或維護地鐵系統。隨著我們的經濟變得更加數字化，這些角色將變得更加複雜，而不是更簡單。實際上，它們是高度掌握的職業。
如果人工智慧提高了對技能的溢價，我們也必須提高對掌握技能的認可方式。但除了認可之外，我們還必須重新思考如何建立掌握技能。當人工智慧改變工作的執行方式時，我們也必須重新思考如何傳輸技能。
如今，我們的許多行業轉型地圖（ITMs）指導部門增長和勞動力發展。但它們在很大程度上是為前人工智慧時代設計的。更新這些框架可能是有價值的，以明確說明人工智慧如何改變學徒制和在職培訓途徑。這個問題以前在這個議會中提出過，值得重新關注。
特別是，我們應該考慮是否需要在ITMs之外建立行業培訓連續性地圖，以確保即使人工智慧承擔更多常規任務，我們也繼續維持深度熟練的人類工人的強大管道，特別是在掌握技能、判斷力和動手專業知識無法被替代的角色中。
如今，這在熟練的技行業中不那麼明顯。這就是為什麼我在上次供應委員會中也提議了一個國家技工大師認證框架，以認可進展、獎勵深厚技能並整合人工智慧能力。
如果我們做對了，人工智慧不會掏空中等技能工作，而是會提升它們。這也將進一步幫助提升我們技術教育學院（ITEs）和理工學院的熟練畢業生。
尊敬的議長先生，鑑於人工智慧的巨大機遇和風險，作為增長引擎的人工智慧需要有利於工人和公民的合理政策。阿拉伯聯合大公國和芬蘭等國家已經採納了協調一致的國家人工智慧戰略，結合政府領導、企業採用和勞動力發展。
在新加坡，我們已經採取了重要步驟。2026年預算宣佈了由總理勞倫斯·黃領導的國家人工智慧委員會的建立，以及通過企業創新計劃為企業提供激勵措施，並通過SkillsFuture和TechSkills Accelerator為工人提供支援。
我想提供一些建議，以更有效地裝備工人和企業，確保沒有工人掉隊。
首先，我們必須使人工智慧成為生活的常態。我們必須超越培訓途徑和時間受限的人工智慧工具訪問。雖然我歡迎通過政府和NTUC倡議提供臨時訪問，但我們必須思考在訂閱初始階段之後會發生什麼。
許多更強大的人工智慧工具需要持續的訂閱。隨著時間推移，這可能會在能夠定期使用這些工具的人和無法使用的人之間造成分裂。如果不加以解決，我們冒著在人工智慧'有者'和'無者'之間創造新形式不平等的風險。
由於對人工智慧工具的訪問直接影響生產力、學習和收入潛力，不平等的訪問將轉化為不平等的結果。因此，我們應該考慮如何確保持續和可負擔的訪問，特別是對低收入工人、自由職業者、技工和小企業。
可能的方法包括基礎級別的補貼性訪問，類似於數字連線，與行業夥伴的分層或基於群體的定價，通過社群中心、圖書館和培訓中心的共享訪問，以及確保接收人工智慧支援的僱主也向工人提供訪問許可權。
尊敬的議長先生，如果人工智慧是包容性增長的力量，獲取權不能是特權，必須被廣泛共享。在數字時代，獲取人工智慧可能會變得像獲取網際網路一樣基本。我們必須確保沒有新加坡人因價格原因被排除在那個未來之外。
推動採用的一種方式是讓政府成為有用人工智慧工具的'第一客戶'，並緩解與人工智慧轉型相關的擔憂。啟用人工智慧的系統可以為利用政府服務的公民提供更快和更實用的回應。隨著人工智慧系統持續向新加坡人提供有用的結果和快速建議，對人工智慧和人工智慧採用的信心將增長。
在這一背景下，我也想承認政府為支援工人適應以人工智慧塑造的經濟、保持就業為中心成果並緩解工作焦慮所做的努力。
因此，需要承認，例如，在M³和焦點地區4（FA4）框架下，馬來/穆斯林社群的就業成果已經在規模和有針對性支援下實現。在2022年至2025年間，新加坡勞動力部和NTUC的就業和就業能力研究所（e2i）協助了超過29,000名馬來/穆斯林求職者，其中超過19,000人成功獲得就業。
與此同時，通過M³和FA4下的社群途徑，超過6,000名求職者獲得了接觸，其中超過500人成功就業，包括需要更持續支援的求職者。這反映了我們國家就業系統的廣度和我們社群為中心的干預的深度。基於此，FA4工作流將加強其對支援工人適應以人工智慧塑造的經濟的關注，保持就業為中心成果。
我很高興NTUC將通過NTUC的e2i與MENDAKI緊密合作，加強職業轉換，特別是對於從校園過渡到職業的年輕成人。這些年輕成人可能在對工作關聯性和人工智慧驅動的崗位流失的高度不確定中進入勞動力市場。這包括通過與高等教育機構（IHL）合作並整合e2i的職業服務、工作匹配和僱主網路來加強早期職業途徑，使新畢業生為變化中的勞動力市場做好更充分的準備。
對於服務不足的馬來/穆斯林工人，作為初步步驟，MENDAKI和NTUC的e2i聯合試點了社群環境中的Langkah Digital AI研討會，並計劃在今年進一步擴大規模；我計劃參加其中一些研討會。總的來說，這些有意的就業相關干預將有助於確保人工智慧的生產力收益不會導致無就業增長，而是為新加坡人賦能，跨越生命各階段，適應、保持就業能力並充滿信心地進步，得到NTUC、其夥伴和更廣泛的勞動力運動的支援。因此，我鼓勵我們的社群利用這些舉措進行技能提升。
第二，我們必須關注支援人工智慧的基礎設施。新加坡已經投資於強大的數字基礎設施。諸如Singpass這樣的系統已經允許安全交易，包括具有法律約束力的流程，例如持久授權書。下一步是通過應用程式程式設計介面（APIs）增強互操作性，以便更多服務可以無縫整合。
當服務整合時，人工智慧可以顯著提高效率和使用者體驗。與此同時，我們必須對資料共享框架和安全港進行調整，以便資料可以被負責任地使用而不會阻礙創新。
第三，我們必須確保僱主重新設計工作流程以有意義地嵌入人工智慧。議員Mark Lee已經提到了這一點。單獨的培訓是不夠的。雖然企業勞動力轉型計劃提供了有用的支援，但它傾向於覆蓋已經傾向於轉變的大公司。我們需要採取進一步的行動。
一種可能性是推出"人工智慧雙語"認證，針對僱主而不僅僅是工人和求職者。雖然它可以是自願"選擇加入"的基礎，比如BCA的綠色建築認證計劃或TAFEP的公平僱用徽章，但該認證可以與某些其他利益或配額關聯，以激勵公司主動參與。與現有的自願計劃一樣，這可以與激勵措施聯絡起來，以鼓勵更廣泛的參與。
第四，我們必須支援傳統僱主的員工、技工、低工資工人和平臺工人。人工智慧可以充當個人助手，增強生產力和收入。例如，技工可以使用人工智慧工具生成報價單、發票和客戶回應；低工資工人可以使用人工智慧進行日程安排和財務規劃；平臺工人可以最佳化跨平臺的路線和工作，增加自主權。政府是否有可能投資於此類工具並提供時間有限的訪問權，以便這些工人能夠體驗其實際好處？
最後，我們必須認識到並非所有工人都有相同的能力適應人工智慧。時間限制、照顧責任和人生階段挑戰影響參與培訓。我們應該向前推進，促進靈活學習；將培訓融入工作；並加強跨部門流動性。這確保我們的勞動力保持靈活、流動和包容。尊敬的議長先生，請允許我現在用馬來語發言。
(用馬來文)：【請參閱《方言演講》。】我們希望每一位工人都能獲得他們所需的支援以提升技能，不被人工智慧的持續進步所甩在身後。通過與馬來/穆斯林工人的互動，許多人將人工智慧視為一個機遇，但也擔憂他們無法跟上這種快速的技術進步。
這種擔憂是有效的。人工智慧正在改變我們的工作方式。特別是PMET開始提出疑問：我的技能仍然相關嗎？我能適應這些變化嗎？根本問題是：這個人工智慧經濟中還有我的位置嗎，還是我會被甩在身後？
作為經濟韌性委員會的聯合主席，與萬里扎爾博士合作，我們的目標明確——通過以開放和準備的態度擁抱人工智慧轉變，來建設和加強我們社群的經濟韌性。我們希望的增長能創造機會並賦能工人，而不是取代他們。
我們將評估經濟戰略審查委員會將闡述的經濟轉變的影響，識別新的行業和增長機會，以及瞭解我們如何能夠鼓勵馬來/穆斯林社群更廣泛地參與這些領域。同時，我們正在制定有針對性的戰略，以加強社群在經濟轉變倡議中的參與，確保參與能夠在所有階層——從青年到專業人士和企業家——中得到深化。
在馬來/穆斯林社群中，這項工作已經開始，並必須通過M³（現更名為M³+）來加強。讓我分享一些由馬來/穆斯林機構所進行的努力的例子，以提升我們社群的人工智慧素養。
我們看到了令人鼓舞的倡議。在新加坡伊斯蘭宗教理事會（MUIS）及清真寺，人工智慧素養計劃幫助社群理解負責任地使用技術。IftaSG倡議在法特瓦研究中運用人工智慧。鼓勵深思熟慮的人工智慧整合的計劃也已向伊斯蘭宗教教師提供，以實現更豐富和更有意義的伊斯蘭學習體驗。
在新加坡伊斯蘭學者和宗教教師協會（PERGAS），人工智慧培訓通過多項計劃為伊斯蘭宗教教師配備數字技能，這些計劃包括《多樣性驅動的伊斯蘭學者技能提升》、《人工智慧加速器挑戰賽》和《伊斯蘭學者企業家人工智慧》。
在穆斯林專業人士協會，針對在職專業人士，《學習圈：關於生成式人工智慧的一切》計劃將在本月舉行。該計劃將探討生成式人工智慧如何重塑我們的工作方式，以及專業人士如何開始在其日常角色中有意義地應用它。
在MENDAKI，MENDAKI成就計劃（MAP）現在使用Khanmigo和KiteSense Luminee等人工智慧工具來提高來自弱勢背景學生的學習成果。這體現了確保技術進步成為促進社群社會流動和包容性增長催化劑的承諾。
Abdul Kadir bin Abdul Rahman先生是一位科學和數學領域的資深教育工作者，他很好地展現瞭如何將三十年的深厚經驗與當代創新相結合。作為MAP計劃的培訓師，他是運用人工智慧技術改善學習質量的堅定倡導者。他指出一個重要的變化是學生提問意願的提高——這促進了更具互動性和支援性的學習環境。
此外，MENDAKI的Langkah Digital倡議提供人工智慧就緒工作坊、實踐培訓和技能提升計劃，幫助個人理解和應用人工智慧於日常生活和工作中。MENDAKI與新加坡科技設計大學（SUTD）等機構建立合作伙伴關係，開啟了人工智慧、設計和應用型學習的機會——使我們的社群不僅僅是技術消費者，而是能夠掌握技術的主體。
這表明人工智慧可以成為社會流動的催化劑——如果我們確保獲取機會和資源被廣泛共享的話。但並非每個人都從同一起點出發。有些人擁有機會、支援性環境和學習時間。有些人則面臨約束——無論是時間、家庭責任還是信心不足。這就是為什麼我們的方法必須具有包容性。培訓必須易於獲取、切合實際且有用，以便每個個人都有機會與這些變化相適應並取得進展。最終，人工智慧經濟中的成功不僅由技術衡量，而是由我們確保每個公民都能夠充滿信心和希望地向前發展的程度來衡量。
（英文）：議長先生，人工智慧將帶來破壞和機遇。如果管理得當，它可以提高生產力，為所有人擴大機會。但如果管理不善，它可能加劇不平等。
我們必須確保人工智慧推動的不僅僅是增長，而是包容性增長。如果我們處理得當，人工智慧不會分裂我們的勞動力，而是會加強它。
通過這樣做，我們將更新這份人工智慧驅動增長的契約。在這份契約中，每個新加坡人，無論是用程式碼工作還是用雙手工作，都在我們的經濟中有一席之地、有一個角色、有一個未來。我衷心支援這項議案，議長先生。
主席：楊女士。
下午1時09分 楊婉玲議員(榜鵝)：議長先生，不久前，我在榜鵝的一個紅綠燈處時，一輛自動駕駛穿梭車駛過。我不是唯一觀看的人。在我身旁，司機和行人都抬起了頭——既有好奇心，也有一種更加安靜的情感。一種低沉的嗡鳴般的焦慮，伴隨著一絲敬畏。在他們的眼睛後面，有一個非常人性的問題：這對我意味著什麼？
那個場景一直留在我心裡。人工智慧和自動駕駛技術正在改變我們工作、生活和娛樂的方式，速度比人類有史以來看到的任何技術都要快。因此，本議院今天必須回答的問題不是這些技術如何工作，而是我們在做什麼——具體地、有意地——以確保我們的工人的生活和生計不被落下。
議長先生，這就是這項動議的真正含義。我想對此講話，不僅僅是給出保證，而是提出一個計劃。一個為我們工人的計劃，一個為我們工會成員的計劃，一個為坐在旁聽席上支援我們這項動議的兄弟姐妹們的計劃。
這種轉變已經在靜靜地發生，無處不在。在樟宜機場，自動行李拖車在航站樓之間運送行李。在濱海堤壩服務路，自動清掃機清掃樹葉和垃圾。在巴西班讓碼頭，無人自動導引車在貨場之間運輸集裝箱。我們的第一批創造收益的自動駕駛巴士服務計劃在今年下半年在兩條路線上執行。
技能提升理所當然地處於最前沿。但工作重新設計同樣至關重要，以重新設計現有工作以適應新現實，建立新的工作型別，並在人工智慧改變工作週期時支援工人的過渡。要使工作重新設計真正產生效果，它必須是一個真正的自下而上的努力，以工人和他們真實的工作流程為中心。讓我詳細說明。
首先，有意地諮詢工人，以真正瞭解他們的工作。作為全國運輸工人工會(NTWU)執行秘書，我親眼看到過三方合作有效運作時的樣子。我們的管理合作伙伴SBS運輸、SMRT和其他公司一直在為我們的巴士車長和技術人員做準備，以應對人工智慧、電動車(EV)和自動駕駛車(AV)的到來。僱主提供技能提升和培訓。政府在這一過程中支援公司和工人。工會則做我們最擅長的事：在工作現場認真傾聽工人的真實需求。
正是這種地面層面的傾聽提出了我們本來會忽視的問題。即使我們朝著到2030年我們巴士車隊的50%成為電動車的目標前進，我們的巴士車長指出電動車培訓存在重要缺陷。與使用鏡子的傳統巴士不同，電動車使用數字監視器，我們的車長告訴我們關於時間延遲、眩光、眼疲勞，更嚴重的情況下還有噁心。他們要求更長的培訓和準備時間。工會與我們的三方合作伙伴一起推動了這一點，問題得到了解決。
議長先生，這是任何顧問報告都不會提出的反饋。但它直接影響巴士設計、駕駛員安全和乘客體驗。工人比任何人都更瞭解他們的工作。這是我們必須繼續利用的資源。
為了應對我們的第一批自動駕駛巴士服務，全國運輸工人工會(NTWU)去年調查了約500名巴士車長和技術人員。三分之一的人對自動駕駛車會影響他們工作表示擔憂——工作保障是最大的擔憂，其次是對降薪的恐懼。這不足為奇。這些是全球運輸工人所共有的情感。然而，我們調查物件中有三分之一仍然相信司機將繼續發揮重要作用。
因此，我們進行了更深入的調查。我們與巴士車長坐在一起，要求他們向我們講述他們一天的工作，不是他們的職位描述說的，而是他們實際做的。
他們告訴我們的話顛覆了我們的假設。從紙面上看，我們假設駕駛是巴士車長工作的核心，也許佔他們任務的80%。我們的車長告訴我們，這更接近20%。其他80%——幫助老年乘客安全上車、管理車廂擁擠、緩解衝突局面、給乘客指路、成為車上冷靜和令人放心的存在，甚至告訴乘客只能在坐著時進行歌唱表演——這些是任何自動駕駛車都無法替代的深層人性化責任。
這有深遠的影響。如果我們按照關於巴士車長角色的紙面假設行動，我們會誤判工作規模、技能要求和薪酬結構——為工人創造不公平的結果，同時為組織帶來人力資源(HR)規劃災難。正確確定職位描述不是一種官僚行為。這是所有工作重新設計的基礎。
議長先生，工會和三方就業委員會將繼續進行地面工作。但我們不能獨自做這件事，如果我們認真對待大規模的工作重新設計的話。我呼籲政府為此提供適當的資源：資助對實際工作角色和工作流程進行系統研究和製圖，以便工作重新設計建立在實地真實而不僅僅是假設的基礎上。
議長先生，我的第二點是：要使人工智慧轉型成功，工人和客戶必須處於關於人工智慧能為工人和企業帶來什麼的重新想象過程的中心。不是事後諮詢。不是被告知已經做出的決定。從一開始就要處於中心。
AI將改變我們所知的工作。但它確切地會落在哪裡——哪些任務、哪些角色、哪些行業——沒有人能完全預測。這正是為什麼重新想象的過程如此重要。我們不能等到塵埃落定。我們必須與工人們一起建設、準備，是的，大膽地夢想一個由人工智慧驅動的工作場所會是什麼樣子。我從訪問正在進行轉型的公司中學到的最重要經驗是：當你早期且真誠地讓工人參與時，他們不會抵制變化。他們會推動它。
議長先生，讓我給你介紹信任中心（Trusted Hub）。一家新加坡中小企業，經營25年，從事資料處理業務，在其核心，這實際上正是人工智慧所在。回到2001年，信任中心處理來自公眾成員的政府提交；堆堆疊疊的紙張；影印機、傳真機、印表機。快進到2026年，同樣的業務，或多或少相同的客戶，但工作方式完全不同。人工智慧現在處理大部分資料，減輕了他們員工的負擔。
當我訪問時印象深刻的不是技術。是人。因為信任中心將他們的工人作為利益相關者——而不是乘客——納入了重新想象的過程，他們的大多數員工已經提升自己來程式設計AI代理，為公司創造企業和創新價值。公司中最年長的AI代理程式設計師？一位60多歲的紳士。自學的。當你不低估你的工人時，這就是會發生的。
我以前在這個議院談過FairPrice在榜鵝海濱購物中心的未來商店，它在國際貿易展上被推薦為超市未來模型，並作為技術如何能使我們工人的工作更好、更容易、更安全的生活展示。它用人們可以親身體驗和看到的證據而不是言語來驅散恐懼。
但是讓它成功的不是人工智慧。是過程。工人和工會塑造和設計了系統，而不是繼承它。正因為如此，員工不僅接受了變化，他們擁有了它。
我想要的是更多這樣的。明天的商店、明天的公交車樞紐、明天的餐廳、明天的診所。活的試驗檯，允許重新想象不僅在公司內部發生，而且在整個叢集和我們的社群中發生，以便有關工作場所人工智慧的對話可以公開、坦誠和富有想象力地進行，而不是恐懼。
很像我的榜鵝居民看著我們的自主穿梭車滑過，有一種低調的焦慮，是的，但絕對夾雜著一絲敬畏。
雖然明天工作場所的形狀仍在形成中，但有一點是清楚的。將工人和工作流程放在轉型中心不是可選的，這是方法。實踐中是什麼樣子的？
這是Chye Thiam Maintenance公司向自願接受掃地機器人培訓的工人提供200美元培訓津貼，使轉型成為工人選擇的東西，而不是對他們做的事情。
這是Grab與工會合作，評估AV穿梭車安全駕駛員是否能夠承受持續警惕的整個8小時班次，因為工人福利是設計的一部分，而不是事後考慮。
這是一位英國企業家開始稱他的AI機器人為AI員工，以提醒自己和他的團隊，AI不是關於取代人，而是關於改變角色。
這些不是宏大的姿態。它們是小的、有意的、但非常重要的行為，正常化了工作場所中的AI，並使其成為工人可以看到自己在其中興旺而不是被取代的東西。這涉及負責任的僱主、進步的員工，當然還有支援和培養的政府。這是三方制，這是為什麼三方就業理事會對於組織如此重要，從一開始就定下正確的基調，關於AI如何在日常公司生活中被嵌入和獎勵。
議長先生，這是對AI取代的無根據恐懼的真正答案——不是保證，而是證據。證據表明，當工人被視為共同創造者時，轉型更快，採納更強，結果對每個人都更好。
三方就業理事會有利於在每個層級推動這項重新想象工作，通過公司培訓委員會（CTC）、叢集層級的CTC女王蜜蜂、跨行業的行業AI提升計劃。我們的女王蜜蜂可以帶上他們的承包商生態系統，就像FairPrice在他們的未來商店中所做的那樣。工會將做我們最擅長的：與工人和管理部門一起走過前面的路是什麼樣子，以及沿途出現了什麼新角色。
但議長先生，這真的需要投資和意圖。我呼籲政府為行業的行業AI提升計劃提供資金——零售、物流、醫療——具有與已經開始指導我們公共運輸AV路線圖相同的故意性。工人應該知道的不僅是AI即將到來，而且是下一個試驗檯在哪裡，新工作會是什麼樣子，以及如何到達那裡。清晰度不是奢侈品。對於站在那個十字路口的工人來說，這是一切。
議長先生，我的第三點是這樣的：即使是管理最好的AI過渡也會看到工作消失，一些職業發現他們所做的任務被AI接管。這是誠實的事實。我們不應該用樂觀主義來掩蓋它。因此，過渡支援必須是真實的、必須是及時的，必須到達最需要它的人。
我們欠我們的工人一個在他們跌得太遠之前抓住他們並儘快讓他們回到一份好工作的系統。那個系統必須從工作重新設計開始。不是事後考慮，而是作為第一道防線。如果我們很好並及早重新設計工作，我們減少了需要被抓住的工人數量。最好的過渡支援是使懸崖從一開始就更短的支援。
這就是為什麼我們傳送給企業的訊號如此重要。AI贈款必須與強制性工作重新設計要求相關聯，生產力增益與工人成果相關聯。如果這些企業無法留住我們的工人，這些公司應該被要求提前通知政府他們無法留住的人員，以便這些失業工人可以得到e2i和我們新成立的三方就業理事會的協助。這將向工人保證新加坡的AI過渡不會導致無就業增長，我們將過渡時間保持到新的好工作儘可能短。
議長先生，我為我們總理在2026年預算中關於AV過渡將被仔細管理、與平臺工人協會和我們的駕駛員進行密切協商的保證感到欣慰。作為全國計程車協會和全國私人僱車協會的顧問，我想直接談論這個問題。用普通話，請議長。
（用普通話）：[請參閱本地語言演講。]我們的計程車和私人僱車駕駛員已經在競爭激烈且燃料成本很高的環境中航行。看到AV在榜鵝運營，聽到自主公交車試點的訊息，他們不禁心中充滿了一種平靜的、未言明的擔憂。他們不要求我們停止技術進步——但他們需要的不僅僅是保證。他們需要清晰的方向感。
自主車輛的地理圍欄將如何逐步擴充套件？時間表是什麼？新的角色，如遠端操作員和安全主管，正在出現。我希望願意的駕駛員將獲得支援並獲得培訓機會，以便他們可以過渡到這些新職位。對於那些還不能進行這種過渡的人，我也希望新成立的技能和勞動力發展局將花時間更仔細地瞭解他們的需求——因為他們不是一個同質群體，不能被視為一體。
這不僅僅是一個涉及平臺駕駛員的問題。我們的技術人員和工匠用他們的雙手保持新加坡運轉，但在關於AI的對話中，他們常常是無形的。他們的貢獻長期以來一直沒有得到足夠的認可。
我很高興人力資源部（MOM）已經開始在這個領域推動努力，從電氣貿易開始。我們必須繼續努力為工匠建立更有希望和更受尊重的職業道路，同時利用AI增強他們的能力——而不是取代他們的判斷。
（用英文）：議長先生，最容易受到AI破壞的工人往往是緩衝最少的——儲蓄最少、靈活性最少、等待系統跟上他們的時間最少。這就是為什麼我們的反應必須是全面的三方制。僱主和平臺夥伴必須隨著他們的商業模式演變而加倍努力，而不是退縮。這意味著繼續參與過渡支援、共享培訓成本、覆蓋機會成本，以及在就業和就業後途徑中支援工人。
工會將做我們一直做的事情，走進現場、傾聽，並與我們的工人一起塑造生計機會，我們將繼續'jaga rumah'——通過關注其他司法管轄區正在做的事情，從中國關於AI替換本身不是解僱理由的網際網路法院裁定到加州對AV上人類安全操作員的要求。這些是世界各地確定邊界所在的訊號。新加坡必須從中學習，在必要時，走在它們前面。政府必須圍繞需要它的人設計過渡支援，而不是圍繞行政上方便的東西。這是我們必須堅持的標準。
議長先生，我將以三項對政府的呼籲作為結論。
給時間。工作重新設計不能倉促。它需要與工人坐在一起，瞭解他們的工作真正是什麼，而不僅僅是工作描述說的，當第一個答案被證明不完整時回到現場，就像我們在公交車長那樣做的。公司需要得到支援，而不僅僅是被推入其中。
在重新想象方面給予幫助。大多數公司，尤其是我們的中小企業，無法單獨做到這一點。我期待政府提供實際的便利、框架和資金，使工作重新設計成為可能。我期待我們的領先企業、我們的女王蜜蜂，站出來、分享已經完成的工作，並將他們的行業帶進來。停留在一個公司內的轉型是隻有一半完成的轉型。
讓NTUC成為聯絡紐帶。我們的工會、我們的e2i、我們的三方就業理事會——我們已經在現場、在公司中、通過我們的CTC，與工人和僱主每天坐在一起。我們有需要許多年才能建立的信任。我們的勞動運動已經為成為這一過渡的結締組織做好了準備，將失業工人與重新設計的角色相匹配，提倡公平待遇，並讓每個人（包括我們自己）負起責任。SWDA和我們的機構必須以此為基礎。
議長先生，我回想起那個在榜鵝滑過的自主穿梭車。看著它的工人們不是要求我們停止它。他們要求我們確保當它向前發展時，他們也向前發展。這是我們欠他們的答案。不僅僅是一個承諾；一個計劃。
我相信，實現沒有失業增長的AI轉變是可能的。這不是因為技術會自動解決，而是因為我們會這樣做。如果我們妥善諮詢工人，讓他們參與重新想象他們的工作，並通過確實能惠及最需要幫助的人的轉型支援來支援這一點，我們就能做到。我支援這項動議。
主席：吉亞姆先生。
下午1時28分 傑拉爾德·詹燕松先生（阿裕尼）：我宣告我的利益衝突——我是一家為培訓提供商提供軟體的公司的所有人和董事。
議長先生，我們面臨對勞動力的結構性威脅。數十年來，新加坡的經濟模式建立在這樣的前提之上：高學歷和高技能的勞動力將掌握繁榮未來的鑰匙，併成為抵禦經濟風暴的緩衝。然而，我們現在正處於一個範式轉變的時期，人工智慧不僅在增強人類能力，而且在許多方面正在取代它。與過去經濟週期不同——那時的動盪可以作為創意破壞的一個插曲而被消除——人工智慧有望成為我們經濟和社會關係發生根本轉變的先兆。進一步推進這一概念，它甚至會影響政府在個人與社會之間進行調解所發揮的角色。
今天，我們必須認識到勞動力經濟力量的本質正在改變。即使生產率飆升，如果我們不解決這個問題，也將導致固化的下層和中產階級失去經濟代理權。這種關注在Jasmine Sun為《紐約時報》撰寫的評論文章中得到表達，她指出了舊金山共識——人們越來越認識到高度暴露於AI職業的年輕工人招聘已經在下降。她提醒我們存在形成永久性下層階級的風險，其中技術的收益集中在少數人手中。
並非所有證據都指向災難。美國國家經濟研究局2025年的工作論文發現，暴露於較高AI的任務確實經歷了勞動力需求的減少。然而，迄今為止，整體就業影響一直很溫和，因為生產率的增長抵消了一些流離失所。同樣，麻省理工學院的Danielle Li和斯坦福大學的Erik Brynjolfsson在《季度經濟學雜誌》上發表的研究發現，生成式AI工具使工人生產率提高了近15%，其中經驗較少的工人收益最大，從而表明AI可以是階梯，而不僅僅是活板門。
應該注意的是，這些研究檢查了早期和控制的部署。隨著代理AI同時在整個行業擴充套件，分配後果可能比早期生產率研究所建議的更嚴重和更迅速。
我們無法確定新加坡走的是哪條軌跡。風險的不對稱要求我們為更難的情景做準備，而不是更容易的情景。
這種關切得到了AI革命的建築師的共鳴。2021年，OpenAI執行長Sam Altman在其部落格文章《一切的摩爾定律》中預測，AI會將權力從勞動力轉向資本，聲稱如果公共政策不相應適應，大多數人會比今天的情況更糟。至關重要的是，Altman並非宿命論者。他主張主動重新分配AI驅動的財富，包括給予公民經濟中的股權，可以使這成為廣泛繁榮的轉變。
同樣，Anthropic執行長Dario Amodei指出，民主的健康基於普通人通過創造經濟價值而擁有槓桿，這是他在2024年的文章《愛的機器》中表達的觀點。
這種槓桿的侵蝕是一個令人深感關切的前景，需要大膽而結構性的政策回應。新加坡處於獨特的位置來領導這一回應，並捕捉AI為我們人民所呈現的真實經濟機會。作為一個擁有高學歷勞動力、強有力機構和資本充足的主權財富基金的小型開放經濟體，與許多大國相比，我們擁有迅速和結構性地採取行動的工具。
但這個機會視窗不會無限期地保持開放。雖然成本套利使離岸外包具有吸引力，但AI可能會侵蝕這一優勢——不是通過將這些工作帶回來，而是通過使小型新加坡專業人士團隊能夠完成曾經需要數百名離岸工人的工作。機會不在於傳統意義上的回岸，而在於在國內集中更高價值的協調和監督角色，其中信任、機構質量和接近決策者最為重要。
AI的平等化潛力超越了白領工作。一名英語困難的藍領工人可以用母語口述，讓AI即時呈現為專業文件，使他們能夠專注於自己的工藝而不是語法。AI應該是一個平等化工具，提升技術大師，而不是分層我們勞動力的楔子。AI工具還可以通過使小型超高效團隊創造巨大價值並擴充套件，以最少的人力實現全球範圍，來推動新一代本地初創企業。
因此，新加坡必須走在這一轉變的前沿，同時確保利益流向所有公民。這需要受過培訓、擁有技能並善於利用AI工具和創新的工人和企業家，並賦予他們的員工也能這樣做的能力。
我們目前重新培訓新加坡人的努力常常受到低效用外部培訓計劃陷阱的阻礙，這些計劃產生的認證在AI驅動的經濟中缺乏現實世界的價值。這些計劃讓培訓提供商獲利，同時讓工人掌握經濟價值甚微的技能。這種錯位存在創造兩速經濟的風險，其中資本所有者和技術整合公司將那些陷入傳統就業慢速道的人留下，導致社會凝聚力的根本侵蝕，並增加長期結構性失業的風險。
為了解決這一問題，我提議建立一個國家AI公平基金。該基金是維護我們社會契約完整性的必要保障。這是一個戰略性的盈餘轉移，從從AI中獲益巨大的企業轉向新加坡人，以促進我們的集體穩定。
在我解釋基金用途後不久，我將詳細說明精確的融資機制。我建議基金按兩個不同的支柱進行組織。
第一個是社會紅利，收入作為直接支付分配給每個成年新加坡公民。我提議初始公民紅利為每個成年公民500美元，隨著基金貢獻的增長而向上擴充套件。這在設計上是溫和的。它不是為了替代收入，而是為了提供一個切實的訊號，即每個新加坡人都在我們的共同未來中擁有所有權。
根據我們目前的公民人口，這將每年花費約15億美元——或少於去年預算盈餘的10%——併為每個新加坡家庭提供有意義的回報。這將充當社會底線，確保國家數字繁榮的收益為所有人提供切實的心理平安和尊嚴。
隨著工作性質的演變，這一紅利將為家庭提供額外的緩衝。它還使新加坡能夠獲得AI的全面生產率優勢，而不會過度加劇社會不平等。
可以辯稱，社群發展委員會（CDC）代金券已經這樣做，但那些完全是自由裁量的。我提議的社會紅利是一個結構性權利——收據的函式，而不是財政情緒在那一刻碰巧是什麼。這個區別對於規劃其未來的家庭來說意義重大。
基金的另一部分將專用於掌握基金，這將是一個僱主主導的在職培訓（OJT）模式，將培訓從課堂移出，進入每個企業。
我提議掌握基金為任何進入或轉換到AI增強角色的新加坡公民提供掌握學徒工資，覆蓋其總薪資的50%，以中位工資為上限，為期六個月。這獎勵了工人的適應努力，同時降低了企業在這個不穩定市場中聘用、培訓和留住人才的障礙。
認識到許多中小企業缺乏設計結構化OJT計劃的能力，我建議基金也融資一個專家OJT顧問庫。這些在OJT設計方面經驗豐富的顧問將在企業之間輪換，以構建根據每個企業具體需求定製的OJT藍圖。這將幫助中小企業填補人才缺口，同時也解決為市場新進入者建立培訓和學徒計劃新階梯的需求。
此外，我建議向僱主提供導師學分，以補償高階員工用於結構化指導的時間，將我們的工作場所變成真正的掌握學院，並確保技能與經濟的實際需求保持相關。
掌握基金應提供給在新加坡成立和以新加坡為基地的所有商業實體和社會，包括微型企業。應密切監控資金的使用，以確保其確實有助於每個企業內的AI掌握。我估計掌握基金的年成本約為14.2億美元。
讓我列出融資細節。
第一個來源是對年度利潤超過1億美元的企業的公司所得稅率邊際提高兩個百分點。通過關注這些公司，我們從那些最能通過AI而不是人數驅動增長的公司獲取自動化盈餘。無論是全球科技公司還是傳統巨頭，這些企業都處於將收入與勞動力脫鉤的最前沿。這項稅收增加將每年產生約15億美元，確保創紀錄效率的收益被回收到國家AI公平基金中，造福所有新加坡人。
第二個來源是針對性地增加我們投資回報的利用。我提議將預算中納入的最大淨投資回報從50%提高到52.5%，額外的2.5%直接流入基金。根據目前的估計，這將每年籌集約14.5億美元。
我們的主權財富實體GIC和淡馬錫一直是進入AI領域的早期行動者，投資於Anthropic等基礎公司，並與Microsoft、BlackRock和Nvidia一起承諾數十億美元用於AI基礎設施合作。由於這些全球投資從全球範圍內勞動力自動化中獲利，我們只需將這些收益的一小部分回收到我們自己的勞動力中。
轉移2.5%並非激進的要求。這確保了我們的儲備不僅能提供財務穩定性，還能為每位新加坡人提供長期的經濟自主權。
當我們展望未來時，我們不能簡單地假設失業工人會像在以往的技術革命中那樣順利地轉向新角色。蒸汽機沒有取代人類的判斷力，但人工智慧可能會這樣做。這正是為什麼被動技能再培訓是不夠的，為什麼需要社會紅利的財務保障。轉向創業、護理工作、技工、體育和藝術等自動化程度較低的工作的工人不僅需要培訓，還需要時間和安全保障來做出這種轉變。
當然，會出現我們還無法想象的新工作，但我們必須建立一個足夠強大的系統來支援我們的人民，即使這些工作的出現速度比我們希望的要慢，或分佈不夠均勻。國家人工智慧平等基金為新加坡人提供了財務緩衝，使他們能夠自信地進行這些轉變。
在今年的撥款委員會辯論期間，我提議了青年工資信用計劃——針對聘用年輕新加坡工人的僱主的有針對性的工資補貼。國家人工智慧平等基金將這一邏輯擴充套件到對所有在人工智慧轉變和其他技術破壞中應對的新加坡人的更廣泛的長期框架。
議長先生，國家人工智慧平等基金是我們數字時代社會契約的更新。我們不能讓人工智慧成為分裂我們社會的楔子。相反，我們必須利用它成為我們國家歷史上最偉大的平等化力量。通過建立社會紅利和精通基金，我們給予每位新加坡人在我們數字繁榮中的直接利益和保持領先的資源。
讓我們的目標是確保隨著機器變得更加有能力，我們的人民變得更加安全。通過現在採取行動，我們能夠確保技術進步為每位新加坡人的尊嚴和經濟自主權服務。先生，我支援這項動議。
議長先生：Poh Li San女士。
下午1時43分 Poh Li San女士（Sembawang West選區）：議長先生，在這次會議中，實際上在過去幾次會議中，很少有演講沒有提及人工智慧時代。政府也談了很多全球中斷將如何影響我們的娛樂、工作和生活。
在政策制定中，總是存在二元選擇：乘上人工智慧浪潮或被淹沒並落後。這是霍布森的選擇，答案是顯而易見的。但我們不能針對虛假的對手來論證。新加坡是一個小型、開放的、數字化連線的經濟體。人工智慧將成為經濟生活的事實。
政府表示將增長我們的經濟、支援我們的企業和照顧我們的工人。但是，部委的政策制定與實地實施之間存在差異。
在街道層面，人工智慧轉變看起來令人生畏、昂貴，對許多中年工人來說，是一個焦慮和困惑的地方。這是第一波變化，可能也是最艱難的。政府、企業、工會和工人必須共同努力度過這一波浪潮。
在轉變期間，一些工作會消失，但新的工作也會出現。如果由市場決定，只有最強大、最適合和最有能力的受益。在一個經濟叢林法則不受阻礙地運作的城市中，讓我們坦誠地說，人工智慧轉變將使一些人受益，但不是所有人。這種增長不會導致所有人的生活更加美好和繁榮。
但這不是市場的工作。這是我們的工作；我們在這個議會中的所有人都必須將市場的力量引向我們的目標：為新加坡人創造更多高價值工作並重新培訓失業工人，但要以滿足企業利潤需求的方式進行，以便我們的經濟能夠繼續長期增長。我們過去通過滿足兩方利益的解決方案建立了新加坡，我們必須在未來再次這樣做。
讓我具體談論人工智慧可以造福我們工人的兩種方式。
我們過去說人民是新加坡的唯一資源。我們現在是一個超高齡社會，總生育率為0.87。我們的資源池在萎縮。人力資源現在是大多數企業的關鍵瓶頸和成本驅動力，特別是中小企業，它們僱用了70%的本地勞動力。如果企業關閉，更多工人將失去工作，即使那些不受人工智慧威脅的人也是如此。
在過去的十年中，我們的企業面臨日益惡化的勞動力短缺。2025年Manpower集團最新的人才短缺調查顯示，亞太地區近五分之四的僱主在尋找熟練人才方面存在困難，77%報告了困難。特別是，許多新加坡人不能或不願做的工作由非新加坡人完成。
但我們對外國工人的依賴是有限的，包括政治和社會限制。我們需要人工智慧驅動的機器人來替代外國工人，那些沒有新加坡人願意做或能做的工作。例如，建築、海事和航空部門的重型工作，這些工作暴露於變暖氣候的惡劣環境中。這對我們來說將是改變遊戲規則的。
下一個前沿是物理生成人工智慧或具體化人工智慧。最近，生成人工智慧技術已與物理系統整合，使機器能夠與現實世界互動並適應現實世界。它使機器人能夠通過模擬和將智慧從數字模型轉移到現實硬體來學習複雜任務，如操縱和導航。簡單地說，能夠思考甚至感知的機器人和類人機器可以被部署在非結構化的和動態的環境中，以協助或甚至替代人類工人。
最近幾個月，Dexterity AI、Figure AI和Unitree Robotics等公司已經展示了它們在人工智慧驅動的機器人和類人機器在專業角色中的能力。
與像ChatGPT這樣在網際網路上搜索以訓練其模型的生成人工智慧工具不同，物理生成人工智慧工具需要在特定背景的環境中為它們將要做的角色和任務進行訓練。隨著時間的推移，這些物理人工智慧能力將成熟並可供面臨人力短缺的企業使用。這些人工智慧驅動的機器人可以幫助我們的企業克服人力約束、降低成本並提高盈利能力。
物理人工智慧機器人擅長重複任務，但無法替代每一個角色。重新設計為人機混合團隊的工作和流程將成為新常態。老年人和女性可以加入改變後的勞動力——重複的、重型任務由機器人完成，複雜的監督角色由人類執行。這將是一個新的自由和賦權的模式，今天無法想象但在不久的將來將成為現實。
新加坡人可以被培養成機器人的監督者。將為年輕工程師和技術人員建立新的高價值工作角色，如這些人工智慧機器人的設計、建造和維護。
如果他們願意，更多新加坡人可以在更晚的年齡退休，因為他們的角色將變得不那麼體力要求。老年人和女性可以加入以前由那些有更強身體能力的人主導的行業。而且機器人也不攜帶任何社會包袱。
議長先生，向人工智慧驅動的機器人的過渡是我的工作領域，這是我們正在努力實現的願景——為企業解決真實的問題並提升工人的生活質量。我強烈地認為，我們的人工智慧轉變應該專注於為我們的產業定製物理生成人工智慧解決方案，以幫助每位新加坡人在這個賦權的旅程上。
總理Lawrence Wong為我們的國家人工智慧戰略概述了四個關鍵支柱。特別是，在先進製造和運輸連線部門，人工智慧驅動的機器人將確實是力量倍增器。
我們為這種轉變做好準備了嗎？還沒有。我們為此已經做好了充分準備，但我們必須快速行動。我建議採取以下六個步驟。
第一，工會應該預測哪些型別的工作和部門處於風險中，以及可能失業的工人數量。
第二，人力部應該為受影響的工人資助再培訓，以為他們準備其他角色或其他產業。
第三，工人也應該站出來，學習新技能並對新工作機會持開放態度。
第四，教育部（MOE）和IHLs應該重新設計學術課程，以使學生遠離已經被人工智慧佔領的領域。所有IHLs學生都應該學習與其學科相關的人工智慧工具。
五，貿易及工業部應吸引更多世界一流的物理生成式人工智慧公司在新加坡設立總部，並吸引人才進行研究與開發（R&D）。
六，企業應願意與人工智慧公司合作，以自動化和重新設計工作流程、改革工作職位並建立人機混合團隊。議長先生，我想用馬來語分享幾點意見。
（用馬來語）：【請參考民族語言演講。】議長先生，新加坡的人工智慧轉型必須由政府、企業、工會和工人共同管理，以避免收益不均。物理人工智慧可以緩解勞動力短缺，並將工作重塑為人機混合團隊。通過工作重新設計、員工技能提升和教育改革，工人可以進入更高價值的職位，而企業可以通過更高的生產力實現增長。
為了支援這一轉型，政府的支援和監管至關重要，以資助工人再培訓、創造高價值工作、吸引領先的人工智慧企業，並確保人工智慧被道德地用於廣泛的社會利益。
（用英語）：在不久的將來，一個新的人工智慧生態系統將出現。技術公司創造人工智慧解決方案，企業擁有它們，工人們利用它們。
但政府必須制定規則。人工智慧必須被用作善的力量，而不是用於犯罪和有害的活動。建立人工智慧使用的倫理規範將決定我們的社會是從人工智慧中受益，還是被其奴役。
但關於人工智慧還有一個更深層的道德問題。人工智慧是人造的；它本身沒有內在的善，沒有固有的價值。我們在這個議院的議員有責任引導人工智慧在市場中的使用方向，不僅要禁止犯罪行為，更要推進公平、善良和正義的使用。我們必須確保人工智慧轉型不僅僅創造增長，而是創造就業、惠及工人、加強企業並提升社群。議長先生，我也想用普通話來總結我的觀點。
（用普通話）：【請參考民族語言演講。】議長先生，新加坡的人工智慧轉變需要政府、企業、工會和工人的協調努力，以確保企業和員工都能受益，並支援和協助受影響的人。這包括提供再培訓和擴大教育範圍，創造高價值工作，使更多新加坡人能夠儘快適應這些變化。
物理AI（結合生成式AI的機器人）可以通過在建築、航空和海事等行業承擔艱苦危險的工作來緩解勞動力短缺，同時將工作角色轉變為人機協作模式。
我想提議幾項關鍵步驟來支援AI轉型：
首先，企業應該自動化並重新設計工作；其次，工會應該標記處於風險中的職位；第三，工人應該繼續學習；第四，政府應該制定清晰的AI規則。國家轉型不應僅僅為了追求更高的經濟增長，而應確保社會各階層都能受益。
（以英文發言）議長先生，我起身發言的動議要求我們確認AI轉型'不能導致無就業增長'。這個'不能'不是經驗預測，也不是空洞的修辭。這是政治決心。
自由市場中的AI可能或可能不會成為我們人民自由和賦能的新模式。是我們的決心讓它成為如此。議長先生，我支援動議。[掌聲。]
議長：低安德烈先生。
下午1時57分 低武楊·安德烈先生（非選區議員）：議長先生，提交給這個議院的動議呼籲一項不會落下新加坡工人的AI轉型。總理、勞工首長、整個政府在過去幾個月都說過同樣的話；這就是他們的意圖。我今天下午想要審視的是，我們擁有的政策框架是否足以滿足我們被要求確認的承諾。
議長先生，企業做出的每一次人工智慧部署，本質上是一項選擇。企業可以使用人工智慧讓現有員工變得更有能力、更富有生產力、比以前更有價值，或者用人工智慧完全不需要這些工人。經濟學的簡記法將此稱為增強與自動化的對比：增強是指人工智慧與工作者並肩工作，自動化是指人工智慧替代他們。
斯坦福大學經濟學家埃裡克·布林約爾松是人工智慧與勞動力市場領域的主要學術聲音之一，他提出了一個令人信服的論點：在沒有有意政策調向相反方向的自由市場中，激勵機制系統性地傾向於自動化。企業發現部署人工智慧來替代工人比重新培訓他們更容易、更便宜。稅收制度、勞動力市場機構、資本成本結構都傾斜了競爭環境。即使增強隨著時間的推移會創造更多的總價值、更多的好工作、更廣泛的繁榮和更公平的收益分配，但無指導系統的預設軌跡是自動化。
政府選擇並宣佈的方向是增強。今天擺在我們面前的議案假定了增強。昨天本議院的勞工首長用他自己的話表達了同樣的承諾——不是人工智慧代替工人，而是人工智慧為工人工作。
這一哲學方向在議會兩邊已經確定。實質問題是我們的政策框架是否與之相符。
目前有三個地方的架構校準不當。三個地方，在當今，制度儘管作出了相反的承諾，仍在允許自動化。
勞工部長昨天表示，人工智慧也在重塑高階專業人士、經理和行政人員(PME)職位，例如醫生、律師和會計師。總理表達了同樣的觀點——人工智慧將影響新加坡的專業人士、經理和技術人員(PMETs)，他們花費多年建立了專業職業，現在被告知腳下的地面在移動。
上週在五一勞動節集會上，總理陳振聲表示，'我們可能無法保護每一份工作，但我們將保護每一個工人。'問題是政府選擇的工具——SkillsFuture求職者支援計劃——是否能夠兌現這一承諾。總理將求職者支援計劃稱為'新加坡方式'，是相比工人黨(WP)首選的冗餘保險更實用、更符合新加坡特色的替代方案。這是對新加坡傳統的倒讀。
議長先生，勞工部長昨天在議會中表示，過渡期間的財政支援不是福利，而是對工人成果的投資。按照這個標準，這一傳統一直以來就建立在正是這種投資之上。中央公積金(CPF)、MediShield Life、MediSave，這些都是全民繳費計劃，在生活的重大風險發生時支付。每一項都覆蓋每一個工人，因為它所保險的風險可能會影響每一個工人。這就是新加坡方式。
求職者支援計劃並非建立在這一傳統基礎之上。它是一項以稅收為資金來源的補助金，以裁員前收入為門檻，在設計上更接近於經濟援助而非風險保險。按目前的配置，它在六個月內以遞減的月度分期支付最多6,000元，首月為1,500元，最後三個月每月為750元，並且僅向被裁員前月收入為5,000元或以下的工作者提供。
勞動部長昨日在本議院承認該上限排除了面臨人工智慧時代相同失業風險的PMEs，並提議將符合條件的上限提高到接近PMEs中位總收入水平。
如果採納該提案，這是向工人黨長期主張的方向邁進。但Ng先生的提案只是移動這條線，我們的提案是將其移除。提高上限讓更多工作者進入該計劃，但不改變該計劃對他們的作用。對於確實符合上限資格的工作者，遞減支付傳達了一個資訊：從高金額開始並逐步減少的支付不是一個缺陷。這是一個倒計時。倒計時會推動工作者接受第一個工作機會，而非正確的選擇。
人力部的資料本身告訴我們為什麼這很重要。在去年最後一個季度中被裁員的居民中，43.6%的PMETs在六個月內沒有找到新工作。這是求職者支援計劃無法應對的群體。而在那些在六個月內找到工作的人中，大約十人中有四人的工資低於以前。所以，他們接受了可用的工作，而非他們經驗相稱的工作。我們大多數人都經歷過，你在職業階梯上爬得越高，找到下一份工作的時間就越長。
議長先生，一個被六個月倒計時推入他們不想要的低薪工作的PMET已經經歷了政府框架本應阻止的自動化結果，雖然附帶了微小的緩衝。提高上限只是擴大了該群體，但不會縮短倒計時。
議長先生，工人黨對裁員保險計劃的提案建立在新加坡實際的傳統基礎之上。我們支付最後一次所領工資的40%，沒有收入上限，也沒有遞減機制。它由僱主和僱員繳款資助，採用與中央公積金相同的模式，並覆蓋所有為該計劃繳款的工作者，包括勞動部長已認定為風險最大的專業人士，因為它所保險的風險不止於$5,000、$7,600或議會可能設定的任何其他上限。
總理說我們必須保護每一位工作者。政府選擇的工具沒有做到。工人黨的做到了。
議長先生，當一家公司考慮進行重大人工智慧部署時，它面臨兩條道路：一條道路是保留現有員工並對他們進行再培訓，使其與人工智慧並肩工作；另一條道路則是裁員，精簡運營，引進規模更小、懂人工智慧的勞動力。前者是增強，後者是自動化。
但我們的稅法對處於決策關頭的公司說了什麼？目前的框架獎勵活動。它獎勵人工智慧資本支出。它獎勵培訓支出。這些都是值得獎勵的事情，但目前的框架所沒有做到的是獎勵這一選擇本身。
裁減現有員工並培訓較少新員工的公司獲得與保留和重新培訓現有員工的公司相同的財政待遇。購買人工智慧來替代工人的公司獲得與購買人工智慧來增強工人的公司相同的財政待遇。
稅法在這個選擇關口沉默無語。
正如布朗森先前所指出的，在選擇關口沉默無語並非在後果上保持中立。當稅法不主動獎勵員工保留時，潛在的經濟學會使公司傾向於裁員。畢竟，勞動力是資產負債表上最昂貴的一項，而勞動力成本是永久性的，這與一次性培訓成本不同。在市場無干預的情況下，公司會選擇裁員。
昨天，黃先生在本議院為CTC框架進行了辯護，將其作為將企業轉型與工人進步掛鉤的機制，並提議通過新成立的三方就業委員會進行擴充套件。
CTC在專案層面為與其合作的公司運作，附帶補助金資金。擴大其覆蓋範圍會擴大補助金模式，但不會改變每家公司都在其中運作的更廣泛財政體系結構，無論其是否在CTC計劃內。正是這種更廣泛的財政體系結構塑造了首席財務官在決策點上的財務決策。
二月在本議院，我提議推出再培訓稅收抵免，這是一種僅適用於能夠證明他們已將現有員工保留在人工智慧增強職位中而非裁員的公司的扣除。正是這一缺失的有條件部分將在公司必須做出決策的確切時刻為其提供財政訊號。這項再培訓稅收抵免將獎勵積極的選擇，而不僅僅是投資人工智慧。
本議案的第四部分確認經濟進步必須保持包容性。這是關於分配的承諾，而不僅僅是增長。我的同事Gerald Giam提議建立一個國家人工智慧公平基金來從結構上實現這一承諾。我今天提議的工具是任何再分配機制（包括Giam先生的）需要運作的診斷工具。因為增強戰略的第三個條件要真正實現就是驗證。
議長先生，增強最終是一個可測試的主張。它做出了一個預測，即在人工智慧與工人並肩部署的部門中，工資將跟隨這些工人幫助創造的生產率增長。如果這個預測成立，政府所採納的框架就正如宣傳的那樣得到了實現。如果這些部門的生產率提高了，但工資並未隨之增長，那麼所實現的就不是增強，無論我們用什麼語言來描述它。
目前，我們很少有機制，也很少有系統的方式來區分究竟發生了哪種情況。
政府正在四個國家人工智慧使命部門——先進製造、連線、金融和醫療保健進行大規模嚴肅的公共資金投資。公共資金通過CTC補助、新成立的三方就業委員會、技能和勞動力發展局以及各種企業轉型計劃流入這些部門和其他部門。這些是適當的投資，但公共投資產生了相應的公眾問責責任。公共資金流向之處，公眾有權知道產出是什麼以及流向何人。
因此，我要求的是一個有針對性的透明機制，即年度人工智慧收益審計，最初專門針對四個國家人工智慧使命進行範圍界定，向議會報告來自國家支援的人工智慧投資的生產率收益如何在工資和資本回報之間分配。隨著時間推移，其範圍和覆蓋面可以擴大。
二月在我的預算演講中，我將其定義為分配問題。今天，隨著本議案在議院面前要求我們確認經濟進步必須保持包容性，我再次提議將其視為更根本的東西。人工智慧收益審計是議會可用的最直接工具，以測試政府選擇的增強方向是否真正得到實現。如果收益與工人分享，審計將說明這一點，框架將有證據支援其主張。如果沒有，我們將在差距變成鴻溝之前、在本議案從政策陳述變成希望陳述之前瞭解到這一點。
議長先生，增強和自動化之間的選擇不是一天做出的。它每天都由我們執行的計劃架構、我們維持的稅法和我們選擇收集的資料所做出。無論議院今天說什麼，那個架構將繼續代表我們做出選擇。
目前，我的立場是該架構將工人推向第一份可得到的工作，而不是合適的工作。我們的稅法對於在重新培訓工人和保留他們之間的選擇關口沒有任何肯定的表述，我們也沒有建立任何機制來判斷公共人工智慧投資的收益是否惠及了那些該投資以其名義進行的人。這就是我支援本議案的原因。我敦促政府為其提供所需的架構，以便我們能夠確保沒有工人被遺忘。感謝您，議長先生。
議長：哈米德·拉扎克博士。
下午2點10分 哈米德·拉扎克博士（西海岸-裕廊西）：議長先生，我宣告我是一傢俬人骨科診所的業主，該診所已工會化，我也是衛生服務僱主工會（HSEU）的顧問。
我起來支援本議案——無失業增長的人工智慧過渡。人工智慧已經來臨。這不是試點；它已經在成為一個平臺。
先生，我們的問題不是我們是否採用人工智慧。我們會採用。更寬泛的問題是我們是否在增長的同時不遺棄我們的工人。在接下來的十年裡，新加坡應該被評判的不是我們部署人工智慧的速度有多快，而是我們在多大程度上將採用轉化為更好的工作、更好的工資和工作場所更強的信任。
我今天以三個身份發言：作為一名專業人士、作為一位家長和作為傾聽居民意見的議員。
首先，是專業上的焦慮。許多專業人士、經理、執行人員和技術人員（PMETs），他們不害怕技術，他們對不確定性感到不安，因為人工智慧很少替代整個工作。它拆解任務，它壓縮團隊，它改變僱主的招聘物件。當你看不到你的角色如何演變時，焦慮就會增加。
第二，是為人父母的焦慮。今天的父母問非常簡單的問題：我的孩子會有一個公平的開始嗎？入門級工作會是什麼樣子？如果入門級工作萎縮，那麼誰來培訓下一代？
第三，是居民的焦慮，這是最實際的。職業中期的工人擔心失業。照顧者擔心時間。許多人不能只停止工作去培訓或再培訓。他們不是在要求保障。他們只是要求公平的機會和一個他們可以駕馭的體系。
議長先生，我們應該坦率地對待人工智慧。人工智慧是聰明的，但它不是智慧的。它可能產生幻覺，它可能聽起來自信，但仍然是錯誤的。因此，未來不應該是人與人工智慧競爭。它應該是人與人工智慧一起工作，進行判斷、進行驗證和承擔責任。
本議案不僅僅是關於技術。它涉及信任、工作重新設計和整個工人旅程。信任將決定這種採用是否會成功。如果工人將人工智慧視為監控，信任就會減弱，最終破裂。當信任破裂時，採用將放緩，收益將不可持續。
總理黃女士談到了保護每一位工人，並擴充套件實際的三方工具，如用於人工智慧過渡的CTC。這是我們應該加倍推進的方向，我在這方面提議四個實際舉措。
首先，技能必須是一條路徑，而不是選單。SkillsFuture是一項重大的國家資產。但在實地，許多工人告訴我這一點。它很有用，但也令人不堪重負。課程太多，徽章太多，訊號太少。
因此，問題現在不僅僅是獲取。它是導航。工人不應該需要滾動數小時來猜測什麼真正對他或他的下一份工作重要。因此，我建議我們按部門、按職位策劃更清晰的人工智慧相關路徑，具有明確的前門和明確的僱主認可。我們可以考慮為那些選擇支援人工智慧驅動增長的優先課程的人提供額外激勵，特別是當有明確的僱主需求時。這可能意味著更高的補助金層級或與結果掛鉤的支援，如完成加面試、實習或根據設計和可行性確定的再部署路徑。
第二，將人工智慧採用與工作重新設計掛鉤。許多議員都談過這一點。如果我們資助採用，我們應該問，任務將如何改變，工人將如何被重新部署，效能測量將如何繼續保持公平？生產率必須體現為對我們的工人更好的工作，並最終體現為更好的工資，而不僅僅是人員數量的減少。
這不是為了懲罰。這是為了務實。這是我們三方夥伴可以通過提供劇本、模板和諮詢支援來幫助的地方，這樣中小企業就不會被丟下獨自應對。
第三，為專業部門（如診所、律師事務所、會計師事務所）提高人工智慧就緒度。許多是小型機構，PMET密集且時間緊張。他們想採用人工智慧，但擔心安全、機密性、責任和信任。
一個已在本議院分享的實用模式已經在醫療保健領域顯現。今年4月，HSEU和GP+合作社簽署了一項協議，為基層診所員工進行人工智慧意識培訓，並幫助基層診所採用技術和重新設計工作流程，由CTC方式和CTC補助支援。我近距離觀察了這一夥伴關係。價值在於使採用在務實、負責任的基礎上進行，並以工作重新設計為錨點，而不僅僅是工具推廣。
我希望我們能將這種基於叢集、CTC風格的方式擴充套件到其他專業部門，包括法律和會計，以便較小的執業機構能夠從不確定性轉向就緒狀態，具有明確的治理標準和工人保護。
對於我們的工人，當需要被識別時，支援應該開始。這意味著更快的工作匹配、適應現實生活日程的模組化培訓和負責任過渡的實際指導。實際上，結構化的指導和明確的下一步減少了焦慮，因為它使工人從等待轉變為行動。議長先生，我現在將用泰米爾語發言。
（用泰米爾語）：【請參閱方言發言。】尊敬的議長，許多人在談論人工智慧時感到擔憂。他們擔心工作可能會消失，技能的價值可能會下降，以及孩子的未來可能會發生什麼。這些擔憂是真實的。問題也很困難，但我們的反應不是害怕。
人工智慧發展迅速，但人類的仁慈、信任、正義感、創意、語言、文化——所有這些都不能完全被任何機器取代。人工智慧可以計算；但人類的聯絡、人類的判斷和人類的責任將始終與我們同在。因此，我們必須將自己從恐懼的道路轉向機會的道路。人工智慧的興起並不意味著人類不再被需要。相反，它更清楚地表明瞭人類必須做什麼真正重要的任務。
在這方面，人文學科很重要；語言技能很重要；文化細微差別很重要；社會理解很重要。醫療保健、護理、教育、社會服務、諮詢和涉及與人直接接觸的工作——這些是人工智慧無法替代的領域。
泰米爾語和泰米爾文化在這一努力中是一種支援；我們的文學培養人類的感情。我們的文化加強了社會責任。這是我們的力量。所以，這是我們必須告訴年輕人的：擁抱人工智慧，但也要發展人類的能力。
學習不僅僅是認證。這是一條道路、一個信念、一個未來計劃。當增長到來時，就業機會必須伴隨而來。變革的支援也必須伴隨而來。所以，不要害怕。讓我們有信心。
（用英語）：議長先生，無失業增長必須意味著一件事。工人能夠感受到的增長，無論是工資、尊嚴還是明確的下一步。
所以，我提出一個治理標準。測試不是我們有多少專案。測試是工人是否能夠快速看到正確的路線，為了正確的工作，在正確的時間。父母是否能對孩子的起點和未來感到自信，以及公民的旅程是否感到無縫。
如果我們保持這個方向，並以三方決心完善交付，新加坡可以有效、負責任和快速地部署人工智慧，同時加強信任和保護尊嚴。基於這些觀察，議長先生，我支援這項動議。
議長：何挺如女士。
下午2時19分 何挺如女士（後港）：議長先生，新加坡對人工智慧的方式經常被國際機構和諮詢公司（如BCG）以及傑出人士（如國際貨幣基金組織（IMF）總裁克里斯塔利娜·格奧爾基耶娃）引用。
我們的技術基礎設施和提升工人技能的舉措是我們計劃如何應對這項新的、快速發展的技術所呈現的中斷和機會的關鍵部分。
然而，我們也必須認識到並採取行動應對另一個不舒適的現實。新加坡是最容易受到人工智慧中斷影響的經濟體之一。國際估計表明，發達經濟體中約60%的工人從事高度暴露於人工智慧的工作。對於新加坡，這個比例似乎明顯更高。因為我們是一個高技能、以服務為導向的樞紐，國際貨幣基金組織的估計表明，我們本地勞動力中約有77%高度暴露於人工智慧中斷，我們的過渡可能比許多其他經濟體更劇烈和更尖銳。
人工智慧如何根本上重新配置我們的勞動力市場？這可以通過三個不同的轉變來理解。
首先，許多現有工作將從內部進行轉變。人工智慧正在或已經接管日常資訊處理任務——起草、摘要、資料提取和標準化分析。管理人員、衛生專業人員和法律專業人員已經在使用人工智慧工具來處理這些型別的任務，從而釋放出用於判斷、複雜問題解決和人際互動的時間。
第二，一些工作將被取代。在經濟學家所說的高暴露、低互補角色中，人工智慧可以自行執行大多數核心任務，保持人類參與的理由更少。文職支援工作者和許多商業和行政助理專業人員，其工作圍繞日常文件、基本處理和標準化客戶查詢，面臨最高風險，他們的職位可能會縮小甚至消失。代理人工智慧技術和模型的進步只是加劇了這一影響。在英國，一些金融機構（如投資銀行）因為人工智慧的自動化能力而重新審視了他們對某些角色新畢業生的招聘。
第三，人工智慧也將創造新的工作和新的需求。我們已經看到對人工智慧工程師、資料科學家和人工智慧產品專家的需求上升，但也看到了金融、醫療保健、物流和教育各領域資料精通專業人員的需求。這些新角色往往提供更高的工資，但僅適用於能夠提供正確的技術和互補人類技能組合的工人。
是的，確實，人工智慧工作轉變已經到來，我們正處於重大中斷之中。然而，影響不會在所有行業均勻分佈，也不會對我們的社會和經濟產生均勻影響。目前，人工智慧中斷在白領工人中最強烈，尤其是入門級職位。與歷史上影響藍領工作的以前技術中斷不同，今天的人工智慧將最多影響認知、白領角色——呼叫中心代理、行政官員或初級業務支援執行官，其工作日圍繞標準流程、常規報告和指令碼響應構建，處於人工智慧可以執行幾乎所有核心任務的角色。
在這樣高暴露、低互補的白領角色中，僱主可以合併職位、放緩招聘或重新設計工作，以確保更少的人期望使用人工智慧做更多工作作為簡單理由。如果我們不解決這個問題，人工智慧的好處將最終只由一小部分工人獲得。
研究表明，生產力和財富收益可能會不成比例地落入那些最有利用人工智慧能力的人身上。高技能工作創造的一個有證明的經濟效應是本地服務需求增加。來自主要科技中心（包括舊金山）的研究表明，每個高階工作都與本地服務部門（如零售和食品服務）中約四個工作的創造相關。
即使這樣的溢位效應產生更多工作，這些工作對脆弱工人的質量和可用性對我來說不太確定。在新加坡，較低工資和日常密集的角色更可能由脆弱工人群體持有，他們也可能面臨自動化造成的更大失業風險。
國際機構（包括國際貨幣基金組織和世界銀行）指出，在沒有政策干預的情況下，人工智慧可能會加劇收入不平等。人工智慧驅動增長的溢位效應對低收入工人的益處程度仍然不確定。我們需要新加坡特定的研究，模型化這些分配影響，並將這些資料公開提供，以通知更針對性的政策反應。
我們還必須記住，新加坡人已經感受到房價上升和基本服務成本更高的壓力。這些壓力是真實的。隨著我們作為一個小的、開放的經濟體，嚴重依賴資本流入，這些壓力已經建立了一段時間。人工智慧的影響會推動進一步的不平等財富積累嗎？因此，問這個問題是公平和迫切的：人工智慧驅動的經濟活動是否會無意中增加日常成本壓力？
除了廣泛的經濟壓力，我們必須轉向這一過渡的人性面。隨著工作繼續被重新塑造，工人繼續升級技能，我們不能將那些面臨系統性障礙的人留下，因為我們的國家朝著人工智慧就緒的未來前進。其中包括殘疾人、婦女、低收入新加坡人，以及年輕畢業生。
人工智慧可能會對殘疾人引入新的歧視形式。由於人工智慧演算法通常通過模式識別進行訓練，它們基於資料集中的常見模式做出判斷。因此，如果使用有偏見的歷史資料來為人工智慧進行培訓，例如招聘流程，人工智慧可能會加強對殘疾人的工作申請和任何其他在歷史上在這個領域代表不足的群體的偏見。
女性工人也面臨被人工智慧邊緣化的風險增加。2024年國際貨幣基金組織關於新加坡勞動力市場的報告發現，女性在人工智慧密集的科學、技術、工程和數學（STEM）角色中代表不足，在具有人工智慧工程技能的工人中也是如此。STEM中的女性佔入門級職位的29%，管理職位的24.4%，但僅佔C級職位的12.12%。總的來說，這意味著她們在被認為是人工智慧安全面的方面代表不足。因此，她們在人工智慧補充高技能工作的地方受益的位置會更差。此外，2026年3月釋出的國際勞工組織資料發現，由女性主導的職業暴露於GenAI風險的可能性幾乎是由男性主導的職業的兩倍，在檢視高自動化風險時顯示了更強的差異。
總的來說，這造成了雙重劣勢。女性工作者不太可能從人工智慧的益處中受益，同時更容易面臨失業風險。簡言之，女性面臨更高的風險，機會也更少。
對於我們的年輕畢業生來說，人工智慧對工作崗位的重塑帶來了新的不確定性。初級工作崗位的流失給Z世代帶來了兩難困境。雖然公司仍在尋求為經驗豐富的職位招聘專業人士，但由於工作崗位被人工智慧吸收，年輕畢業生獲得此類經驗的機會減少了。
2025年，超過20%的畢業生無法獲得全職永久職位，較2023年增加了近5%，超過60%的畢業生稱求職變得困難，年輕畢業生對找到全職工作感到更加焦慮是很自然的。
最近的研究表明，僅僅意識到人工智慧增強或威脅一個人工作的潛力就可能加重職業倦怠，主要是通過加劇工作者的工作不安全感和情感疲憊。
雖然人工智慧通常與白領工作的中斷有關，但易受傷害的工作者和家庭也面臨巨大風險。
對人工智慧工具和培訓的不平等獲取可能會加劇現有的劣勢。那些缺乏家庭資源或有利於學習新技能的家庭環境的人可能會發現自己落後更多。如果不予解決，這將冒著代際不平等加劇的風險。
衡量人工智慧對新加坡超越經濟產出的影響，這一切對新加坡意味著什麼？就在過去的一週，我們見證了《婚姻和生育重啟工作組》的啟動。人工智慧進步對我們的出生率有什麼影響？經濟不安全已經被年輕新加坡人列為延遲或放棄生育的原因，但障礙不僅僅是財務問題。工作不確定性侵蝕了對未來的穩定感和信心，以及有足夠穩固基礎建立家庭和紮根的感覺。如果由人工智慧驅動的中斷加深了這種更廣泛的不安全感，我們可以合理預期對我們本已悲劇性低迷的總生育率的進一步下行壓力。
政府必須採取更有針對性的措施，確保所有工作者，無論其性別、年齡、職業、收入和無障礙需求如何，都為人工智慧造成的中斷做好充分準備，以減輕因裁員而面臨財務壓力的易受傷害工作者的負擔。這將最大限度地減少隨著人工智慧替代變得更加普遍，失業對工作者及其家庭造成的不確定性和代價。
為了確保我們的政策有效，我們需要更多公共資料來衡量人工智慧對我們勞動力市場的中斷。例如，我們如何衡量人工智慧專案的成功。
根據我在今年2月24日提出的議會問題的後續跟進，我注意到人工智慧學徒計劃目前通過三個主要指標進行評估：一是培訓的從業者總人數；二是參加人工智慧學徒計劃的畢業生中從事人工智慧工程相關職位的百分比；三是專案質量的完成和監督。這是一個很好的開始，但它們並沒有告訴我們人工智慧專案和中斷對社會不同群體的影響。這些措施側重於吞吐量而不是公平性。我們需要關於專案完成後兩到三年內人工智慧職位的工資軌跡、工作質量和保留率的資料。我們需要更多資料。
首先，應該更詳細地公開參與者資料檔案。這可以包括以前的職業、培訓前的收入等級、年齡、性別、教育程度和殘疾狀況。這使我們能夠看到參與者來自何處。高接觸、低互補職位，或已有的高互補職位。這將表明易受傷害的群體是否甚至已經開始參與。
其次，更準確地衡量人工智慧在更廣泛勞動力市場中的中斷可以採取暴露互補性對映的形式，從而理解職位是否具有高暴露和低互補性，並建立一個調整框架來跟蹤不同人口群體的失業、工資變化和工作質量。這類資料為政府提供了人工智慧如何影響不同社群的更清晰圖景，以便可以將支援指向最需要的地方。
我現在轉向關於我們的年輕人如何應對人工智慧挑戰的一些想法。
如果人工智慧替代了大量初級職位，年輕工作者可能會發現建立傳統上推進到高階職位所需的基礎經驗的機會減少。我們國家的一個解決方案可能是更好地鼓勵和支援青年創業。這將使他們能夠獨立獲得寶貴的技能，而不是等待被成熟公司錄用。
這種方法建立在已經開啟的大門基礎上。人工智慧通過部署用於建立網站、分析資料、運營營銷甚至自動化後臺辦公任務大大降低了創業障礙。我們有許多初創企業計劃，如贈款和訓練營，但這些舉措是否為新興公司的完整生命週期提供了充分的持續長期支援？
此外，我們的贈款架構仍然以里程碑為重點且受專案約束，鼓勵合規性而非競爭。我們需要一種文化和框架，承認失敗初創企業的價值，或支援創始人主導的網路隨時間推移的發展。借鑑其他創業中心的經驗教訓，哪些因素阻礙了新加坡建立對創業者更有利的可持續生態系統的能力？
首先，我們必須繼續建立可持續的非正式網路，使創業文化自我維持。我們目前的網路通常是基於專案且時間有限的，偏向於短期指導。然而，研究表明，基於相互選擇和親和力的非正式導師關係遠比行政配對更有效。如果導師關係僅與短期贈款相關聯，我們的年輕人可能會錯失從基於信任的指導中獲得的益處，這種指導可以在例如矽谷看到。在矽谷和深圳等領先的創業中心，非正式創始人網路一直是一個關鍵但經常被忽視的成功驅動因素。它們促進了知識共享、供應鏈聯絡以及從錨定公司衍生的新企業。
新加坡可以從中獲益良多。雖然我們有Grab和Block 71等錨定企業，但亞洲開發銀行指出，我們的生態系統落後於他人，因為我們的合作仍然是政策驅動的，而不是有機叢集的。我們如何能夠減少對創始人的行政負擔，以確保他們不會過度忙於達到贈款里程碑，而是專注於建立他們生存所需的市場競爭力以應對人工智慧驅動的中斷。
一種可能是將正式報告限制在贈款結束時，而不是更頻繁地進行，以達到平衡。新加坡還必須更好地利用我們的錨定企業。Grab、Sea和Singtel等公司擁有深厚的技術專長和產業網路儲備，這些儲備在很大程度上仍被鎖定在公司內部。
我們能否使用有針對性的稅收抵免或同行發展專案的共同投資配對，以鼓勵錨定企業為早期階段創始人運營結構化導師和衍生專案？這將使有機網路能夠在現有卓越儲備周圍形成，而不是寄希望於政府贈款週期會這樣做。
私營部門必須發揮主導作用，政府的角色應從召集人和守門人轉變為催化劑。這就是我們如何開始從工業內部發展創業體系的方式。
我們還必須學會重視失敗。新加坡的經濟強調和社會一致性文化使我們有時害怕失敗。經合組織(OECD)2018年的一項研究發現，與任何其他參與國家相比，新加坡學生對失敗的恐懼更大。然而，創業意味著對失敗的寬容。創始人必須在資訊不完整的情況下做出決定，有意義的創新必須有某種失敗的自由作為支撐。我們必須把失敗視為墊腳石而不是汙點，否則我們最終會扼殺我們試圖建立的生態系統，並使我們的年輕人在中斷時代缺乏茁壯成長的能力。
我們可以通過開始我們自己朝著更好的創業空間的過渡，鼓勵實驗並將創業失敗正常化為增長和經驗來實現這一點。
失敗應該是墊腳石，而不是死衚衕。我們在學校內開始了這一過渡，我們必須擺脫完美分數，在學校內進行創業專案，讓學生了解初創企業的內部運作。我們還應該展示失敗的專案，以表彰它們的附加價值。
新加坡目前的破產框架也可以重新審視，以更好地支援創業者。目前，失敗的創始人面臨與任何其他破產者相同的限制，包括旅行禁令、董事資格取消和無自動豁免，無論其失敗是真正風險承擔還是財務不當行為的結果。我們能否探索為真實創業失敗設立的專門途徑，允許創始人更快獲得豁免，更快恢復董事身份，並將其經驗視為有價值的而不是負債？這不是為了使失敗沒有後果，而是為了確保一個誠實的賭注出錯的代價不會永久阻止我們最富進取心的年輕新加坡人再次嘗試。
最後，我也希望我們利用我們在應對人工智慧過渡中積累的經驗，在區域和全球範圍內發揮作用，因為其他經濟體也試圖應對這種中斷。
新加坡擁有強大的地位，我們已經有意決定在全球人工智慧治理議程設定方面發揮領導作用。我們的管理者角色必須超越框架範圍，我們必須發揮自己的作用以解決近期資料反映的人工智慧開發和使用中的全球不平衡。世界銀行2025年資料顯示，儘管僅佔全球人口的17%，但高收入國家佔Notable AI模型的87%、人工智慧初創企業的86%和風險基金資本融資的91%。對易受傷害的群體和全球南方在人工智慧領域嚴重代表不足的情況，有合理的擔憂。作為負責任的世界公民，我們可以發揮自己的作用來解決這個問題。
最近，我們已經開始通過《海獅專案》為東南亞語言量身定製開發人工智慧工具，認識到發展中世界的許多地區冒著被基於西方資料的人工智慧系統拋棄的風險。我們應該以此為基礎，倡導跨東南亞國家聯盟(東盟)的公平人工智慧獲取，向缺乏開發自己框架能力的國家出口我們的治理專長，並確保管理人工智慧的規則不僅反映強勢者的利益，也反映許多人的需求。
這不僅僅是抽象的外交政策和雄心。它對國內的就業有直接影響。新加坡在全球人工智慧生態系統中的地位使我們有能力塑造人工智慧工具在該地區如何被構建、部署和採用。我們應該有意地使用這種槓桿。當我們的研究人員——
議長：何女士，你還有一分鐘。
何婷茹女士：當我們的研究人員開發出能夠跨越東南亞語言的人工智慧系統時，我們就創造了可以部署在我們自己的服務部門、醫院和學校的工具。當我們的公司在人工智慧採用方面領先時，我們就為新技能、新角色和新產業創造了需求，而我們的工人可以接受培訓進入其中。
我們必須確保新加坡人在這些技術的開發中有一席之地，而不僅僅是被動接收在別處做出的決定。我們的全球人工智慧領導力最終是為確保答案是前者的投資。這種方法最終還具有為新加坡創造更多工作和機會的額外好處，這將是真正的涓滴效應。我支援這項動議。
議長先生：葉鴻翁先生。
下午2時39分 Yip Hon Weng議員（Yio Chu Kang）：議長先生，我宣告我在一家全球投資公司工作，從事生態系統勞動力戰略。
近日來，一架飛機一直在舊金山上空盤旋，拉著一條寫著'停止招聘人類'的橫幅。相同的資訊出現在城市各地的廣告牌和公交車站亭上，還伴隨著諸如'AI員工時代已經到來'這樣的標語。這個活動是AI初創公司Artisan的作品。這不僅僅是一個營銷噱頭。它反映了一種擔憂，即工作的未來可能會將人們排除在外，而不是賦能他們。
我起立支援這項議案，因為這不能是新加坡的做法。人工智慧不能成為向工人發出他們可被替代的訊號。在我在淡馬錫的工作中，我看到了技術如何擾亂各個行業，我想預先闡述我的主要論點。要實現不造成失業的增長，企業人工智慧的採用不能僅僅是購買技術。它必須遵循一條經過深思熟慮的路線：我們必須首先建立人工智慧素養；利用這種素養推動工作流程和職位的重新設計；並確保這種改革為我們的工人帶來切實的、共同的成果。
讓我先提出一個重要觀察。在淡馬錫生態系統中,許多公司已在投資AI。工具正在部署,試點專案在增加。但我們看到的真實制約因素不是技術、計算能力或資本。而是勞動力的準備度。我們不缺乏技術。我們缺乏的是轉型。
在我們與20多家淡馬錫投資組合公司合辦的人工智慧素養工作坊中，我們清楚地看到，碎片化的人工智慧素養仍然是主要瓶頸。我們正與首席人力資源官和技術長密切合作，以彌合採納與實際價值創造之間的差距。從根本上說，這是一個技能匹配問題。在技能供應滯後的地方，機會不會消失，它只是轉向其他地方。
我們正在討論的人工智慧轉型具有結構性、全球性，且在加速進行。原本需要數週才能完成的任務現在只需數小時，很快就將只需數分鐘甚至數秒。變化不再以線性方式進行，而是以指數方式進行。在這種背景下，採納已不再是可選的。在新加坡，雖然大企業在與遺留系統和繁重的合規要求作鬥爭，而中小企業則面臨資本和頻寬的嚴重製約，但兩者都面臨同樣的含義：不採納人工智慧的企業將難以保持競爭力。
但如果AI的採納是必要的，破壞是不可避免的。我們必須清楚地認識到，如果這一轉變管理不當，工人將面臨的風險。一方面，隨著AI降低許多工的成本，對這些任務的需求可能會擴大而非收縮。經濟學家將此稱為「傑文斯就業效應」，其中效率不會導致工作減少，反而導致新形式工作的增加。我們之前見過這種情況。ATM減少了日常任務，但擴大了銀行業務。文書處理器提高了產出，並將工作轉向更高價值的崗位。人工智慧可能會遵循同樣的模式。
但實際情況往往呈現K形結果。有經驗且具備AI能力的工作者獲得不成比例的收益，而缺乏這些能力的人，特別是初級工作者，面臨落後的風險。因此，問題不在於AI是否創造增長。而在於這種增長流向誰。真正的風險不在於AI替代工作。而在於它大規模替代機會。一名工作者可能保持就業，但會面臨職業發展放緩和經驗的無聲流失。我們的任務不是否認這種顛覆。我們的任務是對其進行治理。
議長先生，如果我們要有效地治理這種顛覆，辯論必須轉向。僅僅觀察企業是否採納人工智慧是不夠的。我們必須對採納後發生的情況進行問責。問題很簡單。工人在轉型後是否比之前的生活更好？我們必須追問：工作和工作流程是否被重新設計？收益是否得到共享？如果人工智慧提高了產出，但削弱了生計，那不是轉型，而是排斥。我們必須確保公共資金不為此提供補貼。
因此，我請問政府：我們能否為支援計劃建立明確的條件約束？如果公共補助資金正在資助企業轉向，難道不應該明確將這些補助與全國性、以人為中心的評分卡相掛鉤？一份追蹤新增崗位淨數、工作流程重新設計的規模、工資改善、員工保留和技能提升的評分卡。如果我們認真對待無失業增長，我們的採納指標就必須超越簡單計數崗位，轉向衡量職業發展。
從我們在淡馬錫的AI素養工作中，我們瞭解到企業迫切需要關於領導力、工作重新設計、成果衡量、信任和治理的指導。企業無法單獨應對。
這就是我們的勞工運動、NTUC、e2i和工會發揮作用的地方。我們必須授權他們提供這樣的指導，確保工會領導人和企業管理層坐在一起，同時規劃企業的科技路線圖和工人的再培訓計劃。如果公司需要關於工作重新設計和共同成果的指導，我們的三方合作伙伴必須就在現場與他們一起。
最近宣佈的三方就業委員會是及時之舉，但必須積極彌合企業級轉型與工人級成果之間的差距。因為真正的挑戰不是向企業引入人工智慧，而是將工人納入這一轉型。
這將我們帶到這一轉變的中心挑戰：人工智慧不僅僅是一個技術倍增器，而是一個領導力倍增器。
領導者必須真正相信人工智慧的變革潛力，並以身作則使用它，引導其應用，帶動員工跟上。AI可以寫作、設計和最佳化。但它無法行使判斷力、建立信任或在不確定性中引領人們。這一責任仍屬人類。
議長先生，僱主有著關鍵的責任，但如果沒有正確的領導力，採納AI最簡單的方式就是簡單地裁減人員。我們已經看到這種矛盾在全球上演。我們看到亞馬遜在2026年初宣佈裁減16,000個職位，同時進一步倚重AI以提高企業效率。
當人工智慧主要作為削減人員的戰略推出時，它會引發恐懼。當作為能力提升戰略推出時，它會建立信任。這種恐懼的故事長期以來一直在我們的工作場所上演，甚至在擁有人工智慧工具的員工中也是如此。
當生成式人工智慧剛開始流行時，一些員工積極主動地自行探索人工智慧來提高工作效率，但將其保密。他們擔心如果透露新獲得的生產力來源，最終會被裁員或在沒有額外補償的情況下被分配更多工作。當缺乏信任且收益不被分享時，員工隱藏他們的能力，而不是分享他們。
真正的轉型需要領導者採取長期視角。企業必須預期工人過渡到新的工作方式時出現短期生產率滯後，並且必須為實驗創造空間。然而，我們必須承認，許多在這個充滿挑戰的經濟環境中與高運營成本相抗爭的企業沒有充足的時間。
我詢問該部：我們如何能更好地幫助公司應對這些短期時滯，確保員工再培訓所需時間的成本不會成為裁員的藉口？
議長先生、各位，當僱主引領這一轉變時，有一個我們必須清晰認識的結構性轉變——在工作重新設計之前，我們必須進行工作流程重新設計。
我們要求工人改變，但系統保持不變。人工智慧跨越各個職能和領域，重新想象流程如何相連以及價值如何產生。這種對整個過程的重塑必須在各個角色本身被重新設計之前進行。
然而，我們當前的政策方針中存在一個重大缺口。如今，我們的國家支援在很大程度上專注於個人工作重新設計。我們要求工人適應新的工作範圍，但遺留的公司流程保持不變。
因此，AI通常被疊加到過時的工作流程中，其中存在資料孤島和碎片化。生產力停滯，沮喪感上升。員工抵制變革並非因為頑固，而是因為在破碎的工作流程中使用先進的AI工具深感沮喪。
所以，我請問政府和我們的三方夥伴：我們能否擴大支援計劃，明確關注工作流程重新設計？我們如何能為企業提供專業知識和資金，首先重新設想他們的跨職能流程？如果我們理順工作流程，員工自然會看到這項技術的價值，將慣性轉變為適應的熱情。
議長先生，即使有最好的工作流程，擾亂仍然會發生。一些工人將失去工作，一些崗位的變化速度將比預期更快。我們經常指向SkillsFuture和職業轉換計劃等現有措施。但這是一個殘酷的事實——如果這些措施足夠，我們的工人為什麼仍然深感焦慮?
答案是，人工智慧衝擊以前所未有的速度發展，工作者擔心我們的社會保障網無法足夠快地跟上它。
我問政府：我們如何嚴格追蹤現有措施的速度和有效性，特別是對失業工人的財政支援，以確保它們確實能夠快速響應？
對於留任的工人來說，基本的AI能力將不再是競爭優勢。這將成為繼續競爭的必要條件。任務是幫助工人從僅僅是AI使用者轉變為AI指揮者，即那些知道如何精選、引導和驗證AI輸出的工人。
這引出了我對我們勞動力管道的一個關鍵擔憂。如果AI自動化了草稿、總結和初步分析，我們的初級工作會怎樣？如果年輕畢業生無法獲得真正的第一份工作和他們需要的導師指導，他們將永遠無法獲得早期隊伍依靠的基礎經驗來成長，我們將面臨失去未來勞動力的風險。
我問政府：我們如何與僱主和行業領袖合作來保護和重新設計初級職位途徑，使我們的青年能夠發展成為明天AI指揮者所需的專業判斷力？
總之，議長先生，讓我回到我開始的地方。在淡馬錫生態系統中，我們看到公司正在投資AI、試點新工具並推進轉型。但決定性的制約因素不是技術。而是勞動力是否做好了準備。
我與我們淡馬錫公司的技術長分享了這個故事。一家工廠大力投資了新機器。它們更快、更聰明、更高效。但生產力下降了，不是因為技術失敗，而是因為人被落下了。
在另一條線上，該公司做了不同的事情。他們沒有替換工人，而是對他們進行了再培訓。機器操作員成為了資料讀者。技術人員成為了問題解決者。改變的不是裝置。而是使用它的人的能力和信心。不久，故障率下降了。想法來自於車間。曾經害怕變化的工人開始領導它。相同的機器。相同的工廠。但一個非常不同的未來。
這就是教訓。技術可能會設定步伐。但人決定了方向。在AI轉型中，精通而不僅僅是採用，決定了這個方向是否包容性。
這把我們帶到了擺在我們面前的動議。一個不產生失業的AI轉型不僅僅是關於創造工作。它是關於確保工人與技術一起進步，而不是落後。它是關於將生產力轉化為進步。它是關於將創新轉化為共享成果。
如果採用建立能力，那麼精通必須建立優勢。如果工作被重新設計，那麼技能必須深化。如果創造了增長，那麼它必須被廣泛共享。
但這不會自己發生。它需要協調。它需要領導。它需要信任。這就是新加坡具有獨特優勢的地方。我們的三方合作模式，其中政府、僱主和工人一起行動，給了我們不僅能夠對變化做出反應，而且能夠塑造它的能力。這是我們的秘密武器。
當企業投資時，工人必須被裝備好。當工作被重新設計時，工人必須被納入。當發生破壞時，支援必須是可信的。因為這不僅是技術轉型，它是勞動力轉型。建立三方就業理事會是確保這種對齊在實踐中發生的重要一步。
技術會發展。市場會適應。但我們必須明確我們正在建設的未來。它不能是一個說"停止擁有人類"的未來。它必須是一個說"投資於人"的未來。我們的工人是否進步是我們必須一起做出的選擇，有意圖地和果斷地。謝謝你，我支援這項動議。[掌聲。]
議長：Jamus Lim助理教授。
下午2時54分 Jamus Jerome Lim助理教授(盛港)：議長先生，我在今天辯論中的貢獻很直接。我主張，如果我們的目標是保護勞動力免受可能由經濟範圍內擁抱AI導致的失業，那麼我們今天的努力應該主要集中在促進新招聘的政策上，而不是那些專注於減少失業或推動再培訓的政策。
解釋很簡單。證據表明，迄今為止，由於AI導致的當前工人的失業是溫和且侷限的，而新工人的招聘放緩已經明顯，甚至可能會加速。
原則上，AI的後果可能會在勞動力市場的兩端產生影響。它可能會減少企業的招聘動力，從而降低新工作機會的總數，或者它可能會提高公司的招聘頻率或誘使工人離職。
招聘放緩源於當AI工具允許公司廉價有效地替換他們以前需要招聘人工工作者來完成的工作職能時。這在初級職位尤其如此，因為新畢業生缺乏經驗，新員工通常被分配的相對直接的繁瑣工作使這個群體價值較低，更容易被機器替代。
但正如許多觀察者，包括本院的議員已經指出的，這是一個先有雞還是先有蛋的問題。如果我們不讓新工人進入我們的公司，我們肯定不能期望他們獲得必要的經驗和特定工作技能，使他們在職業中期成為有價值的專業人士。
相比之下，當AI工具表明某些角色不再需要時，就會發生失業，因為它們可以由AI很好地複製。以前由人類完成的任務被替換，如果公司內沒有其他地方可以重新分配給這個人，或者這個人被證明太昂貴，那麼他們被解僱。
從積極方面來看，AI可能會為失業工人開闢在其他地方追求職業的新機會，要麼因為他們獲得了在市場上使他們更有價值的AI相關技能，要麼因為他們可能可以自己創業。
但目前，來自全球研究的資訊很清楚。雖然迄今為止幾乎沒有失業的證據，但有充分跡象表明招聘出現了下降。這在暴露於AI的部門中的情況如此，特別是在生成式AI如ChatGPT出現後，隨著代理AI的成熟，這可能會加速。對於職業初期、初級工人來說，前景特別危險。即使被僱用，這類工人也往往獲得較低的薪資。
這背後的原因是直觀的。AI主要替代機械、可重複和明確定義的任務，這些任務主要由初級員工執行。公司仍然看重高階員工的成熟度和經驗，總的來說，寧願削減招聘並重新分配他們原本忠誠的員工，也不願讓他們離職。
先生，這些趨勢在我們的本地勞動力市場中也是可見的。迄今為止，AI對這裡的失業貢獻不大。MOM的最新勞動力市場報告顯示，自2023年以來，總體裁員人數保持穩定，失業率約為2%，自2022年以來變化不大。
在回應關於PMET部門的問題時，人力部高階議會秘書黃成俊也指出，在金融和資訊通訊等受AI影響的部門中，PMET的裁員人數仍然很低，去年最後一個季度僅為960人。此外，黃高階議會秘書也指出，這些部門在同一時期的空缺職位數量很大——大約是其10倍。這可能會被解釋為新招聘的勞動力市場健康穩健，但我對此持謹慎態度。這是因為正如任何求職者都會告訴你的——有一個職位不等於有一份工作。這些工作必須被填補，最好是由正在尋找工作的新加坡人來填補。
這是畫面不太樂觀的地方。最新的畢業生就業調查顯示，在幾乎每個研究領域中，成功找到工作的畢業生比例都有所下降，大約每四個畢業生中就有一個無法獲得全職就業。
在回應提交的議會問題時，部長李智隆指出，下降是由於後疫情招聘激增，我們看到的只是早期趨勢的平均迴歸。
我不太樂觀。根據我的計算，雖然疫前和2020年畢業生的永久就業率平均在70%左右，但在十年前左右，這個數字接近85%，這明顯要好得多。
雖然青年失業率在2020年代比前兩十年要低，但誠然，在世界其他地方，自2022年以來也穩步上升了大約一個百分點。與以往新職位吸收率遵循經濟週期的情況不同，新加坡經濟目前實際上處於擴張狀態。
從需求方面，也有報告稱一些僱主對招聘新員工持謹慎和不情願的態度，儘管幸運的是這些目前似乎是少數。此外，這種悲觀的圖景也掩蓋了更多令人困擾的病態現象。許多新加坡工人可能不得不接受不能充分利用他們技能和才能的機會。這種不匹配在彙總資料中沒有得到很好的反映。
想象一下擁有本地大學高階學位的畢業生，卻仍然被迫從事外賣配送工作；或者在頂尖海外大學學習多年的學生，自回國以來反覆遭到僱主的拒絕。或者有賬單和需要贍養的年幼家庭的有經驗的中期職業專業人士，儘管按照建議進行了技能提升，但在就業市場上連續數月無功而返。我相信議會中的其他議員在我們每週的民眾接待會上會看到類似的案例。事實上，NTUC和MOM最近發表的兩項研究證實了這些例子。
過度資格現象在職業早期人群中最為突出，而非自願不充分就業與自願不充分就業者之間的差距在30歲以下人群中最大。加之全球經濟政策不確定性的大環境，我們本地勞動力市場可能正進入所謂的招聘「大猶豫」時代，這一現象也在世界其他地方被觀察到。
如果我們的目標是，如該動議的標題所表明的那樣，避免無就業增長，那麼自然而然，我們應該優先考慮那些針對勞動力市場招聘的政策。讓我列舉幾個。
首先，我們可以改善激勵機制，鼓勵公司聘用應屆畢業生。正如我在今年供應委員會討論中所分享的那樣，這需要將現有的畢業生行業見習計劃（GRIT）專案擴充套件到全國範圍，成為跨部門的全國實習倡議。年輕工作者可以自由地將其SkillsFuture學分用於與願意接納他們的公司合作的有薪學徒制和實習專案。
企業，特別是中小企業，應該能夠向人力部(MOM)提交關於在職培訓的可信提案，人力部隨後將利用已為企業預留的SkillsFuture企業培訓券和其他補貼計劃來抵消企業錄用這些學員的成本。我的榮幸朋友Gerald Giam也為此目的提議了一個AI精通基金，這與我在這裡提出的建議是互補的。
第二，這樣的短期專案，我指的是六個月到一年，學徒制和實習也應該納入明確的就業路徑，以員工的合理表現為條件，除非由於經濟情況的變化向僱主授予豁免。
這些實習生應當按照《就業法》被視為員工，並獲得同等的法律保護和權益，包括最少年假，但GRIT實習生目前並未獲得這些待遇。
第三，我們可以在高等教育最後一年、學生進入職場之前，加強社交技能培訓的提供，包括溝通、同理心、判斷力、人脈建設和遠見等方面。研究表明，當工作需求不僅要求完成認知任務，還需要人與人工智慧之間進行迭代協作時，人工智慧對員工的補充作用最為明顯。但我們的畢業生往往在學校期間專注於追求學術能力，導致他們在這些介面職能方面的準備嚴重不足。
第四，如果我們確實堅守我們的信念——即希望我們的畢業生專注於獲得能力而非認證，正如教育部(MOE)在支援我們自主大學中可堆積式微憑證途徑方面所明確表示的那樣，以及最近的研究所證實的那樣——那麼我們應該言行一致，在公共部門中取消那些堅持要求文憑或學位的招聘要求，只要能力能夠通過其他方式得到證明。這可以通過一系列微憑證來證明技能，或者候選人在面試階段通過現場演示來證明。
先生，人工智慧是一種通用技術。如同之前所有的通用技術一樣，人工智慧可能會摧毀和創造同樣數量的工作崗位，但當我們面對這一轉變的前沿時，我們必須為那些受此次推出影響最大的人群奠定基礎，至少目前來說，這些人群顯然是我們的年輕入門級員工。我們需要有針對性的政策來幫助克服在招聘他們時的巨大猶豫。這就是我們確保AI的增長承諾不被數百萬失業工作的恐懼所掩蓋的最佳方式。
議長先生：Neo Kok Beng博士。
下午3時06分 Dr Neo Kok Beng（指定議員）：議長先生，我曾經在哈佛肯尼迪學院擔任創新政策訪問教授十年，目前我仍在復旦大學擔任創新管理訪問教授。
人工智慧就是我們真正所說的顛覆性技術，這是我博士研究的課題。這是範式轉變，當你說範式轉變時，意味著它從一個產業，或者說完全從一個產業轉變到其他領域，這實際上會摧毀舊的工作方式。
讓我給你舉個例子。上週，我訪問了Science Park的一家AI初創企業，所以我們討論了在幾個媒體專案上的合作，而我在這個容納20人的房間裡只看到了兩個人。所以，我問執行長，'你們所有的員工在哪裡？'他說，嗯，他們從不來辦公室，但他們在工作。我說，'你什麼時候看到他們？'他說，'嗯，我上週看到他們了，那時網路宕機了，沒有通訊。由於寬頻宕機，他們回來檢查他們的agents，看看是否真的在工作？'所以，我問，'你怎麼知道它們真的在工作？'答案是，'實際上，他們隨時產生程式碼，一直，24/7。'所以，他擁有agents，它們在24/7工作。所以，我說，'你怎麼衡量生產力？'我問了一個問題，'自那以來，你們替代或完成了多少任務或工作，相當於工程師或編碼工作人員，自去年以來？'答案是五。所以，與去年和今年相比，這個人有agents為他工作，他完成了五個人的任務。
我們可以從兩個角度來看：一是這家公司生產力極高；另一個需要看的是，我們少了五個工作崗位。
你會選哪一個？好吧，如果你是CEO，你就知道該怎麼選。如果你是NTUC秘書長，我就不太確定了。
所以，我問了下一個問題。他碰巧正在為擴充套件業務招聘潛在員工，所以我問他，「你對這個正在讀計算機碩士的人怎麼看？他是否達標？」他的回答是，「根據我與他的討論，他的回答表明他在當前技術方面落後兩年。」
落後兩年，我就想，天哪，所以當這個人畢業時，也許明年，他會怎樣呢？技術總是領先他兩年。
因此，這意味著我們是否應該將這個人置於實際工作環境中，基本上就是實習和在職實訓，或者當這個人畢業後，他真的需要加快在當前能力上的進度，因為AI技術發展速度太快。所以，說真的，這是一場海嘯。
就個人而言，我參與了一些小型非政府組織的工作。我們正在進行一個重點專案，用來監測獨自生活的老年人，這樣如果他們發生任何情況，我們都能瞭解。我們實際上在使用人工智慧，因為這些老年人中大多數說方言，所以我們使用小型機器人伴侶，這些伴侶具有理解方言的能力，可以監測他們。這非常好，因為很難讓照顧者去進行監測和探訪這些人。
因此，這是人工智慧真正非常有用的一個方面。機器人AI真的非常有用，這樣我們就可以處理新加坡人不想做或可能不太適合做的工作。
我從事的另一個專案是成像，利用人工智慧進行磁共振成像。我們改日再討論此事。
問題在於工作場所體驗現在變化很大。那麼，我們、員工、現有人員、PMEs或工人是否擁有繼續從事這份工作的技能。我實際上很高興勞工運動推出了CTCs。這是將人工智慧引入工作場所或與企業合作的一個很好的機制。我很高興為此有這樣的補助金。
但問題是我們如何界定能力素質？一些議員談到工作流程重新設計、流程重新設計——但我們如何知道最後他們是否具備能力？工作場所的能力素質水平在哪裡？因此，我認為新機構SWDA應該能夠與勞工運動部門合作來界定這些能力素質。
但這些能力我們還能如何利用？它們是固定於工作場所的嗎；也就是說，一個工作場所、一家公司？還是具有可遷移性的？
那麼，也許我們可以考慮讓專業機構參與，使得每個級別或每個能力級別的資質——無論是可堆積式的還是微憑證——都能獲得行業認證和認可。這樣一來，這個人在其整個職業生涯中都能擁有可轉移的資質。
人工智慧已經是大勢所趨。與大多數員工一樣，起初我——我不會說懷疑，但不願意改變我的工作方式。但現在，我認為我別無選擇。所以過去一年裡，我一直在使用它。甚至我的妻子也要使用AI生成影片和圖片，雖然她是藝術家，喜歡畫畫——但它已經成為她生活的一部分了。
這是一場巨大的變革浪潮，我支援這項動議。正如我在新加坡工程師協會理事會成員的自願工作中曾做過的那樣，我正是那個實際上提議與勞動運動合作建立青年工程師領導力計劃的人，其中包括高階和全球專案。我認為工程機構和專業機構可以與工會合作，對所有這些計劃進行認證，為工人開闢職業發展途徑。
議長先生：Eileen Chong女士。
下午3時15分 Ms Eileen Chong Pei Shan（非選區議員）：謝謝主席。請用中文。
（以中文發言）：【請參閱書面記錄。】議長先生，我同意動議中提出的許多觀點，其中包括確保在人工智慧轉型中沒有任何一名工人被遺漏的承諾。
然而，要實現這一願景，我們還需要分享採用人工智慧帶來的收益，並確保未來的工人能夠繼續保持競爭力，領先於形勢。目前，當採用技術來提高生產力時，往往是僱主受益。我提議升級《靈活工作安排指南》，賦予其法律約束力，這樣可以確保員工也能享受人工智慧生產力的收益。
當人工智慧提高了生產力和工作效率時，我們應該鼓勵工人利用節省下來的時間與家人在一起、休息、參加活動以及建立與他人的關係。
此外，明天的工人——即我們學校中今天的孩子——已經開始在課堂上接觸人工智慧。一些家長對在小學四年級引入人工智慧是否太早提出了疑慮。神經科學家也指出，過早或過度使用人工智慧和技術可能會使學習變得太容易，從而剝奪孩子發展深度學習能力的機會。
處於成長階段的兒童應該學習如何思考、提出問題和做出判斷。因此，我呼籲政府追蹤並定期報告人工智慧在我們學校的採用對不同年齡學生認知發展的影響。
在人工智慧時代，獨立思考和判斷是人工智慧無法替代的技能。這些正是我們應該傳授給下一代的東西，使他們無論世界如何變化都能保持競爭力和韌性。
（以英文發言）：議長先生，我贊同動議中闡述的價值觀：增長必須具有包容性，每名工人都很重要，在人工智慧轉型中任何人都不應被遺漏。確保沒有人被遺漏不僅僅是確保就業增長。這要求我們致力於分享人工智慧驅動的生產力收益。這也要求我們確保未來幾代的工人能夠在一個由人工智慧和其他可能尚未被髮明的技術所定義的時代蓬勃發展。
議長先生，關於人工智慧驅動的經濟轉型的大部分討論都集中在工人需要提升技能並保持相關性的必要性上。雖然這些努力至關重要，但它們沒有回答另一個同樣重要的問題：我們如何確保人工智慧生產力紅利的公平分配？
目前，僱主是預設的受益者。他們從相同人數的更高產出或更少人數的相同產出中獲益。這樣的生產力提高不會自動成為員工的收益。沒有經過刻意的政策設計，它們往往只會是僱主的收益。
與工人分享人工智慧生產力紅利的一種方式是通過時間。在三月份人力部委員會供應辯論期間，我為靈活工作安排被賦予法律約束力提出了案例。要求靈活工作的權利與擁有靈活工作的權利並不相同。我們不應該依賴於將行動負擔放在員工身上的指南，而這些員工最無法承受這種負擔。
我想在今天重申對靈活工作立法的呼籲。隨著我們討論人工智慧轉型如何能使新加坡人受益，這變得更加顯著。由於人工智慧產生了真實的生產力收益，新加坡工人是否會以節省下來的時間而不僅僅是更高產出的形式分享這些收益的問題，市場本身無法回答。這必須是設計出來的。
在實際中，更多的時間意味著什麼呢？它意味著能夠陪伴孩子、做的不僅僅是支付補習課費用的父母。它意味著更少的照顧者必須在工作和需要照顧的家庭成員之間做出選擇。它也意味著休息，真正的休息，在一個61%的員工感到疲憊、39%的工人不願上班的國家。這些不是軟性的結果。這些是人類能力得到補充和維持的條件。
政府和三方夥伴關係會承諾在塑造人工智慧時代政策時將工人福祉與僱主收益和經濟增長並列優先考慮嗎？如果會，我敦促政府首先通過賦予靈活工作安排法律約束力來開始。
當人工智慧使公司生產力更高時，工人應該對人工智慧節省出來的時間擁有有意義的、可執行的權利。時間來休息和從事人工智慧無法複製的那種人類聯絡，而我們的生育率告訴我們這方面我們正在短缺。
議長先生，現在我想轉向對我們長期成功最重要的一群新加坡人。如果我們說每名工人都很重要，那麼我們必須關注還不在勞動力中的工人。我說的是現在坐在我們學校課堂上的孩子，其中有些坐在上面的走廊裡。
現在，我們小學四年級的10歲孩子正在被介紹人工智慧工具。雖然當然有教師監督和護欄措施到位，但我們也應該提出一個更基本的問題。這種早期接觸是在建立他們在人工智慧世界中茁壯成長的能力，還是在建立對人工智慧驅動工具的早期依賴？一些家長已經在詢問。小學四年級太早了嗎？真正的收益是什麼？更重要的是，權衡是什麼？
這些不僅僅是父母的焦慮。神經科學家也提出了嚴肅的問題。在他的著作《數字幻象》中，神經科學家傑瑞德·霍瓦斯博士提出了一個應該讓我們停下來思考的觀點。當技術使思考變得太容易時，學習的深度就消失了。人工智慧是終極的解除安裝工具。它以最少的使用者輸入進行閱讀、寫作和計算。但我們的孩子還不是想要解除安裝和提高生產力的專家。他們是學習者，學習需要付出努力。它需要人工智慧被設計來消除的認知摩擦。
如果我們的孩子開始在10歲時解除安裝他們的思考，他們就不會發展出發現錯誤、提出有意義問題或形成獨立觀點所需的心理肌肉。那麼，我們所說的人工智慧驅動的個性化學習可能會變成定製的舒適。因為它是無摩擦的，所以感覺起來像是進步，但它可能會繞過我們試圖支援的認知發展。
這讓我想起了我今天早些時候提出的一個關切：人工智慧使用中的公平悖論。我很欣賞李燦烈部長關於教育部如何通過家長入口網站主動與家長溝通並分享指導的觀點，指導他們如何在家更好地支援孩子。但並不是新加坡的每個孩子都有一位在家且具有數字素養、有時間根據指導行動、且有能力在課堂外支援孩子學習的家長。來自弱勢背景、獲得的家長指導較少、進行的非螢幕充實活動較少的兒童可能最終會更多依賴人工智慧，而不是更少。
對於回家到一個人手不足的家庭，沒有人來重新引導、提問或監督的孩子來說，人工智慧將始終產生答案，始終減少摩擦，始終使思考變得更容易。這不是賦能。如果人工智慧依賴侵蝕了它旨在補充的認知發展，那麼風險最大的孩子就是我們盡力支援的孩子。
議長先生，我的同事詹姆斯·林副教授在最近的預算辯論中指出，強者和弱者大學生之間的差距不再在於他們在書面作業中提交的內容，而在於他們願意質疑和思考超越指令碼的意願。這些能力是在多年中建立或未建立的。如果我們在10歲時取代這種努力，我們的孩子不能簡單地在大學時下載它。如果做最多無監督人工智慧解除安裝的孩子是那些已經擁有這些能力較少的孩子，那麼我們不是在縮小差距。我們在擴大它——更早。
一份由布魯金斯學會在今年1月釋出的全球研究發現，人工智慧在教育中的最大風險是在關鍵發展年份中取代費力思考。有趣的是，該研究中接受調查的65%的學生將認知發展的破壞列為人工智慧使用的首要風險。孩子們自己能感受到這種差異。
議長先生，我不是建議我們不在學校使用人工智慧工具。我是建議我們遵循證據，而不是炒作。生成式人工智慧作為公眾已經不到四年。我們還沒有關於它如何影響孩子大腦的長期資料。我們不應該讓技術週期的速度超過我們的孩子應得的關懷。
我也欣賞部長今早提供的更新，由A*STAR進行的新加坡縱向早期發展研究（SG LEADS）將擴大以收集資料，幫助我們理解新加坡兒童的人工智慧使用模式以及人工智慧使用如何影響他們的學習和幸福成果。我希望這也將包括在教育中使用人工智慧工具對我們兒童認知發展的影響，包括它們對他們的執行功能和批判性思維、閱讀理解和持續獨立努力能力等技能的影響。我們必須確保通過給他們提供一個將始終相關的工具來為我們的孩子為我們無法預測的世界做準備：一個強大而獨立的思想。
這份動議正確地呼籲新加坡對人工智慧驅動增長的方法以公平、韌性和所有人的機會為基礎。我同意。這就是為什麼我今天談論了人工智慧轉型挑戰我們保護的兩個關鍵事項：時間和思想。
對於今天的工人，我們應該為靈活工作的權利立法，以便生產力收益被收回作為休息、照顧和聯絡的時間。對於我們的孩子，未來的工人，我們必須保護學習所需的認知摩擦，確保我們在培養獨立的思想，能夠在不求助於數字柺杖的情況下解決困難問題。人工智慧轉型不僅僅是經濟事件。它可能是一代人中最重要的機會，讓我們問我們將如何利用技術返還給我們的時間和能力來建立什麼樣的社會。議長先生，我支援這項動議。
議長先生：維克拉姆·奈爾先生。
下午3時26分 維克拉姆·奈爾（新民）：議長先生，我支援這項動議。人工智慧已經在重塑我們的經濟。它正在提高生產力、啟動新的商業模式並加強我們的全球競爭力。對於新加坡來說，這是一個重要機會。但除了這些益處外，還有一個真實的關切——我們如何確保我們的工人不會比他們能適應的速度更快地被取代？
在人類歷史過程中，經濟增長與新工作的創造相關聯。例如，當國家經歷工業化時，隨著產業開放，為各地的人創造了新的工作。當然，經濟增長也可能不僅僅來自勞動力的增加，也來自生產力的增加。這也是應該受歡迎的，因為它為那些正在工作的人創造了更高薪的工作。新加坡已從這兩種趨勢中受益。
在這樣的背景下，對人工智慧的概念關切是簡單的：它會增加生產力這麼多以至於明顯更少的工作將被需要？這的含義是擁有工作的人少得多或那些控制資本的人將獲得更高生產力的所有益處，而大量人將失業。本質上，贏家將獲得更多，失敗者的數量將更大。
如果我們看一下其他國家的發展，我們會看到政府開始做出回應。他們這樣做不是通過阻止技術變革，而是通過採取適度的保護措施，以確保隨著人工智慧的採用增加，工人得到公平對待。
一個我希望討論的領域是人工智慧在招聘流程中的應用。例如，在歐盟，最近通過的《人工智慧法案》認可，用於就業的人工智慧系統，如那些涉及招聘、評估和績效監控的系統，可能會顯著影響勞動者的生計。因此，這些系統被歸類為「高風險」，需要滿足包括偏見測試、透明度披露和人類監督等要求。
這補充了《通用資料保護條例》第22條，根據該條款，個人有權不受僅基於自動化處理所做的決定的約束，該決定對其產生法律效力或嚴重影響其。在就業環境中，這意味著重要的決定（如招聘或解僱）不能僅由演算法做出，而必須有有意義的人類參與。
在紐約市，《自動就業決策工具法案》要求對在招聘或晉升中使用的人工智慧系統進行定期的偏見審計，並在使用此類工具時告知申請人。
雖然這些不是直接防止失業的法律，但它們確實確保與就業相關的決定不會以不透明或無法問責的方式依賴人工智慧做出。它們還促進公平性，並有助於防止自動化決策中的無意歧視。
同時，在德國和法國等國家，勞動法要求僱主在裁員前遵循結構化的程式，包括與員工代表協商、提前通知以及努力對工人進行再培訓或調崗。我們的全國職工總會（NTUC）正在與我們這裡的僱主進行類似的活動。這些要求可能不是人工智慧特有的，但它們有助於確保過渡以結構化和負責任的方式進行管理。
議長先生，這些例子表明，政府不是阻止技術變革，而是認識到立法可以提供防護措施，並確保公司在採用人工智慧時，以充分顧及對工人影響的方式進行。這對於維持僱主和工人之間的信任特別重要。如果工人感到決定是透明地作出的，並且有相應的保護措施到位，他們就更有可能支援而不是抵制新技術的採用。
對於新加坡，我們可以考慮在涉及人工智慧的就業決定中，是否應該對人類監督有更明確的期望。雖然許多僱主已經採納了這樣的做法，但將這一原則正式化可以幫助確保整個行業的一致性。我們還可以探索工人是否應該擁有更明確的透明度權利，包括瞭解何時使用人工智慧系統來評估其工作表現或影響關於其就業決定的權利。
此外，值得考慮是否應該強化對負責任勞動力轉型的期望。當重大技術變革嚴重影響就業時，可以鼓勵僱主提供結構化支援，例如再培訓機會或員工重新部署途徑。
歸根結底，重要的是要認識到，僅有立法是不夠的，必須由健全的機構、積極的僱主和願意適應和學習的工人來加以補充。我們的目標應該是維持一個這樣的系統：企業保持創新和競爭力，工人感到安全並得到支援，機遇隨著時間不斷擴大。
這引導我進入第二點——我們可以為最可能受到人工智慧影響的工作做什麼。人工智慧在執行日常和基於規則的任務方面特別有效。因此，行政工作等領域的職位，以及涉及初級分析的職位，更容易面臨被替代。
然而，問題不僅僅在於這些工作可能會消失。更重要的是，這些職位常常是進入勞動力市場的切入點，為工人提供進步所需的經驗和技能。如果這些機會減少，工人可能會發現隨著時間推移建立職業生涯變得更加困難。
從這個意義上講，風險不僅是失業，還包括職業發展道路的逐步侵蝕。長期來看，這可能導致這樣的局面：個人從入門級職位升遷到更高技能和更高薪酬職位變得越來越困難。
因此，除了一般的再培訓，我們的重點還應該放在促進實際的職業轉變上。應該支援工作者進入相鄰職位，在這些職位中他們現有的技能仍然可以得到應用並進一步發展。這使職業轉變對工作者更加可行，特別是對職業中期的工作者。
同時，公司可以受到鼓勵重新設計工作，使人工智慧補充而非替代人類工作者。例如，雖然日常任務可以自動化，但需要判斷力、溝通能力和問題解決能力的工作應該保留並加強。這樣，人工智慧就成為提高生產力的工具，而不是勞動力的替代品。
在實際操作層面，採用人工智慧的僱主可以被鼓勵及早為受影響的工人識別相鄰崗位，併為其提供結構化的重新部署途徑。這可能包括短期、針對性強的培訓模組或工作重新設計，使工人能夠逐步建立新的勝任能力。
這些措施有助於確保工人不會突然失去工作，而是通過一個有管理的過渡期得到引導，在這個過程中他們的技能和在各自組織內的角色得以保留。工會、政府和僱主可以共同合作為此制定一個框架。
關於創意產業（包括音樂、文學創作和表演），我們應該考慮是否需要立法或進一步的保護，以應對受版權保護的材料被用於人工智慧訓練的問題，以及對這一問題的救濟應該完全由私法處理，還是政府有必要建立一個框架來保護此類材料。這可能包括對個人形象和聲音的權利。如果完全由私法處理，只有資源充足的個人才能採取行動處理此事項，而如果存在這樣的框架，個人藝術家、作家和其他創意工作者可能能夠從這類保護中受益。
新加坡多年來成功應對了許多經濟轉型。每一次，我們都將對變革的開放態度與對社會流動性和共同進步的承諾相結合。
向人工智慧驅動的經濟轉變將是另一項這樣的考驗。這將要求我們在創新與保護之間取得謹慎的平衡。如果我們深思熟慮地應對，我們可以確保人工智慧成為機遇之源，並且增長保持包容性。我支援這項動議。
議長：法德里·法茲先生。
下午3時35分 法德利·法茲議員（阿裕尼）：議長先生，今日議案恰當地認識到AI的變革力量，並肯定AI驅動的經濟增長必須保持包容性。
我的演講包含三個主要觀點。首先，我們必須保護工人的經濟地位，並防止人工智慧的經濟果實僅僅流向擁有人工智慧模型或生產支援人工智慧硬體的人。其次，我們必須確保人工智慧抗禦型就業途徑對新加坡人保持可行性。第三，也是最重要的是，我們必須堅持技術應該服務於人類而不是相反的理念，因為歸根結底，目標不僅僅是沒有失業的增長。它是在不失去我們作為人類和新加坡人身份的情況下的增長。
先生，議案陳述的第二項促請國會強調，新加坡關於人工智慧驅動增長的方針必須以公平、韌性和全民機遇為基礎，而第四項要求國會確認經濟進展必須保持包容性，新加坡不應出現無就業增長。
這些是極其重要的目標，因為如果人工智慧的興起管理不當，它不僅可能代表技術性顛覆，還可能代表勞動力和資本之間的權力重新配置。例如，上月，Meta宣佈將記錄其在美國的每一位員工的按鍵記錄和工作流程。整個工作日期間將不時截圖。所有這些資料將形成用於訓練AI系統的資料集，這些系統有朝一日可能會取代這些員工。其他公司很快可能會效仿。
雖然這在美國正在進行，但我們在新加坡應該問：Meta等公司是否應該被允許在沒有任何明確保障的情況下以這種方式採集員工資料？關於如何採集和使用此類資料，不應該有更強的保護措施嗎？工作者不應該對由其自身資料創造的價值擁有權益嗎？
如果我們未能解決這些問題，我們將冒著渾渾噩噩地走向這樣一個未來的風險：財富和權力集中在少數幾家科技巨頭手中，這些企業越來越多地使用人工智慧來接管之前由其員工完成的工作。這些公司將繼續成為經濟增長的引擎，投資於規模日益擴大的資料中心和效能不斷提升的半導體晶片，這些投資繼續推動國內生產總值的增長。
作為一個國家，新加坡可能通過我們在這些科技公司中的股份以及與其的夥伴關係而獲益。然而，我們必須確保這些利益（主要流向資本所有者）不會以勞動力為代價。這些發展帶來了真實的風險，即被替代的工人可能會看到他們的經濟力量逐步被侵蝕，大量工人將不得不競爭數量更少的低薪工作機會。
我並不是想危言聳聽，但如果不加制約，可能會出現一種數字農奴制——一個制度，其中工人像古時的農奴一樣，不再受土地或封建領主束縛，而是受演算法束縛。我們已經在即時看到這一未來的開端正在展開。平臺工人已經在為黑箱演算法工作，這些演算法對他們的收入和工作時長擁有很大的控制權。隨著人工智慧的發展，許多認知工作和白領工作將越來越多地被自動化，並可能受到演算法的控制。曾經被認為安全的職位可能不再安全。
因此，我們需要加強工人保護框架，涵蓋留任福利以及工人對其在工作場所建立資料的權利等領域。由人工智慧驅動的增長不能以損害工人利益為代價，也不能進一步加劇經濟權力向資本所有者傾斜的現象。
主席先生，該動議的第三項要求議院為工人和企業提供配備和支援，以抓住新機遇，共同進步。儘管AI威脅要自動化和替代許多現有職位，但仍然存在許多難以使用AI自動化的勞動形式。
例如，水管工、電工、空調技術人員、採血員和其他熟練行業已被評估為不太可能被AI取代。這些工作是一個正常運作的社會的基本支柱。然而，長期以來，新加坡的我們低估了這些角色，在經濟和社會方面都是如此。如果我們真正致力於確保增長保持包容性，我們必須糾正這種不平衡。在許多其他發達國家，水管工、垃圾收集員或空調技術人員的工作薪酬足夠高，可以維持中產階級的生活方式。新加坡的情況並非如此。
我們已經做出政策選擇，用薪酬較低的外國工人填補這些職位，而我們的本地工人則被引導進入高薪白領工作。雖然這對我們來說已經執行良好數十年，但隨著生成式AI威脅要減少高薪白領認知職位的數量，這可能不再可持續。
因此，我們必須提高藍領部門的工資並改善職業發展道路，這些部門目前不太吸引人，但也不太容易被AI取代。我們必須通過政策、文化和教育來提升他們的地位，這樣新加坡人就不再會把這樣的工作視為不受歡迎的。
這可能需要困難的權衡。例如，我們應該重新調整某些部門的政策，以確保本地熟練行業的工資有意義地上升，並吸引更多新加坡人來填補這些職位嗎？同時，我們必須更好地利用我們強大的職業教育機構。我們的工業培訓學院和理工學院應該引導更多學生從事專業化的高價值行業。在一個由AI驅動的未來，工作的尊嚴不必僅僅取決於一份工作是否是白領或高科技。我們必須擴大新加坡人認為有吸引力和有意義的工作範圍。
主席先生，AI無疑將決定我們經濟、社會和生活的未來。但我們不應該讓AI來定義我們作為人、作為公民和作為新加坡人的身份。我這樣說是因為擺在我們面前的問題不僅僅是AI會創造還是摧毀工作。更深層的問題是：在一個由AI塑造的時代，我們會成為什麼樣的社會和什麼樣的人？因為主席先生，如果我們不小心，我們可能在經濟上成功，但在更基本的方面削弱自己。
例如，AI創造了一個知識不再稀缺的世界。文本可以被總結，論文可以在幾秒鐘內寫出，方程可以被解決。我之前談過我作為本科生在難懂的文本中苦惱的經歷。這是緩慢的、通常令人沮喪的工作，但正是通過這種奮鬥，我學會了如何思考，如何質疑，以及如何理解世界。
然而，今天，完成任何認知活動所需的努力已經以遠大於計算器取代算盤或打字機取代筆的程度而崩潰。在鼓勵我們的學生利用AI時，我們如何能確保他們繼續學習如何理解觀念、如何闡述論點、如何解決問題，以及如何培養智力獨立性？
主席先生，我重申我的警告——新加坡必須成為一個AI韌性社會，而不是AI依賴的社會。通過不斷將我們的任務外包給AI，我們可能會削弱或破壞我們的創造力、想象力、判斷力，甚至同情心的能力，或者讓這些實踐技能因缺乏使用而萎縮。
這裡的危險是誘惑使用AI作為思考和解決問題的捷徑。一方面，有些人把AI視為某種第二大腦——把記憶、決策，甚至判斷的某些方面外包給ChatGPT或Claude。
誠然，AI工具可以提高我們的思維能力並充當智力輔助工具。然而，在我們和世界之間創造了一層人工中介的情況下，我擔心AI會削弱我們理解世界的能力。我所說的理解世界，是指能夠按照我們自己的條件，通過我們自己的認知努力來解釋、理解和連貫地感知我們周圍的世界的能力。隨著時間的推移，這將涉及反思性的試錯，平衡我們對世界的深思熟慮的解釋和判斷，與世界如何作用於我們。
解釋和判斷是必須通過不斷定期使用來磨練的實踐技能。我們通過行使、測試和挑戰這些技能來開發它們。對世界和應該做什麼做出判斷是一項獨特的人類任務，不應該輕易放棄。
我不反對第二大腦的想法。我的擔憂更具體——對第二大腦的依賴，如果不加以檢查，將削弱第一大腦的平衡性和反應力。
另一方面，我們看到人們與AI伴侶形成情感依附。這些是刺激同情心但不真正互惠的關係。這表明了人們可能失去對現實世界中人類關係的關注的真實風險。如果人們把這些AI伴侶視為與我們周圍他人發展友誼的艱難工作相比的舒適捷徑，我們可能會看到我們社交網路的進一步貧困。
在兩種情況下，危險都是一樣的。我們開始用人工近似替代真實的人類經驗、意義製造和判斷。當這種情況發生時，我們可能會逐漸失去按照我們自己的方式清晰地理解和駕馭世界的能力。主席先生，現在用馬來語。
（用馬來語）：【請參考方言演講。】主席先生，最後一點。有一種傳統的馬來藝術形式是我心中所愛：潘圖——一種有其特定規則的詩歌形式。潘圖由四行組成，有特定的米數和韻律方案。其意象通常源於自然和日常生活場景，以傳達重要的社會價值觀和建議。
不遵循這些結構和約定的潘圖通常不被視為好的潘圖，如果可以稱之為潘圖的話。許多潘圖仍然在馬來人中被牢記，通過口頭傳承代代相傳。簡而言之，潘圖體現了一種傳統——連線今天的馬來人與我們之前的祖先。
因此，我想問：如果我們教學生用AI生成潘圖，而不是發現自己嘗試這些行的樂趣，我們會失去什麼有價值的東西嗎？使用AI生成潘圖的技能一定會轉化為寫一個好潘圖的工藝，或者甚至欣賞這種藝術形式的美學敏感性嗎？當一個文化主題被簡化為AI輸出時，對馬來語言、文化和傳統的長期影響是什麼？
（用英文）：主席先生，最後一點。有一種傳統的馬來藝術形式是我心中所愛——潘圖。這是一種有其特定規則的詩歌形式。由四行組成，潘圖有特定的米數和韻律方案。意象通常源於自然和日常生活場景，以傳達重要的社會價值觀和建議。不遵循這些結構和約定的潘圖通常不被視為好的潘圖，如果可以稱之為潘圖的話。
許多潘圖仍然在馬來人中被牢記，通過口頭傳承代代相傳。簡而言之，潘圖體現了一種傳統，連線今天的馬來人與我們之前的祖先。
因此，我想問：如果我們教學生用AI生成潘圖，而不是自己嘗試這些行的樂趣，我們會失去什麼有價值的東西嗎？更重要的是，使用AI生成潘圖的技能一定會轉化為寫一個好潘圖的工藝，甚至欣賞這種藝術形式的美學敏感性嗎？當一個文化主題被簡化為AI輸出時，對馬來語言、文化和傳統的長期影響是什麼？
主席先生，我不是在建議我們回到AI使用之前的時代。我們必須適應。但我們需要辨識力。我們必須清晰地看待AI能提供什麼和不能提供什麼，並始終問AI服務的目的是什麼，以及它是否適合該目的。我們必須避免被困在一個以AI為中心的凝視中，這樣我們就只能對現實有一個狹隘的、人工中介的理解，以及理解和關聯世界及其周圍人的能力貧困。主席先生，讓我用潘圖來結束我的演講。
（用馬來語）：【請參考方言演講。】在議會中背誦潘圖時；
最好不要使用AI；
馬來人是有文化修養和禮貌的；
詩人的靈感將不會被放棄。
（用英文）：在議會中背誦潘圖時，最好不要使用AI。馬來人是有文化修養和禮貌的，詩人的靈感將不會被放棄。
下午3:51 主席先生：秩序，我們在議院已經待了接近五小時。我建議現在休息。我暫停會議，將在下午4:10重新主持。秩序，秩序。
會議相應暫停
從下午3:51至下午4:10。
下午4時10分繼續會議。
[副議長（謝耀泉先生）主持]
人工智慧（AI）轉型無失業增長（議案）
[(程式文本) 辯論恢復。(程式文本)]
副議長先生：Kenneth Tiong先生。
下午4時10分 Kenneth Tiong Boon Kiat先生（阿裕尼）：副議長先生，我宣告利益關係，我擔任一家開發AI應用程式並提供AI諮詢的公司的董事。
自ChatGPT釋出以來的三年半里，我經歷了兩個令人敬畏和恐懼的時刻。第一個時刻是2022年11月。GPT-3.5可以迭代軟體功能、生成想法、編寫程式碼。五年前，普遍的觀點是每個人都應該學會編碼。如今，編碼能力供應充足且廉價。電腦科學畢業生，甚至來自斯坦福大學等頂尖學校的畢業生，正在苦苦尋找工作。GPT-2是一個生成有趣打油詩的玩具。三年後，它的後繼者消除了整個職業的稀缺性。我們在2023年和2024年曾討論prompt engineering。這個話題也逐漸消退了。
第二個時刻是2025年11月，當時Anthropic釋出了Claude Code，一個與前沿模型配對的可靠AI代理。我可以讓計算機整夜執行，到最後會有工作完成。這與與聊天機器人聊天是一種不同的體驗。聊天機器人與你來回互動，完善你的想法，滿足你的各種改變想法，對你的演講進行批判性測試。代理除非需要澄清，否則就會直接去做事情。它可能有一點偏差，但你提供輸入後，再過五分鐘或50分鐘，它就會回來並解決問題。一個非常聰慧的初級同事。
現在我們有了AI代理、Claude Code、Codex工具，如果我可以借用網路術語的話，這讓我成為了'Claude信徒'。我在自己的工作中使用Claude Code。我可以給它最模糊的規範，它就會返回最精妙的資料工作流或網站佈局。對於一個無論如何都無法自己設計漂亮網站的人來說，這是一種解放。
使用這些工具的過程中蘊含一種遊戲精神，這是每個新加坡人都應該體驗的。這預示了一個令人疲憊的世界——軟體工程師每週工作80小時，同時整夜執行多個AI代理，這樣總有人或機器在時刻工作。求職者，尤其是白領工作的應屆畢業生，申請數百個職位卻沒有收到一次面試。求職入口網站，如LinkedIn，已經成為簡歷的存放處，在那裡，真實的體驗就像向虛空中投射求職申請。
變化的速度讓我們都感到謙卑。我對任何以'人工智慧永遠不會'開頭的論斷都持懷疑態度，因為這些預測的有效期往往只有幾個月。讓我關心的不是目標，而是誰會在過程中被落下，以及我們是否在建立制度來確保沒有人被落下。
我有三項建議。首先，獲取高階AI（尤其是AI代理）的許可權應該是普遍的，而不是受限於課程註冊或工會會員身份。其次，我們必須以對待國家雙邊關係相同的戰略嚴肅性來對待少數幾家開發前沿AI的公司，因為他們在定價、獲取和部署方面的決定如今直接影響我們的生產率邊界，程度與任何貿易協議相同。第三，我們必須通過升級應對AI速度人員替代的冗員處理制度，為工人爭取緩衝時間。
主席先生，我認為獲取高階AI（尤其是AI代理）是一項權利，而不是特權。智慧能力（就其提升意義而言）不應該因財富而產生差異。我每月在這些工具上花費幾百元錢，因為它們改變了遊戲規則。但對於那些無法負擔的人來說，不平等從一開始就被固定了。這是否從一開始就將他們排除在外？
政府部分採納了我的同事Gerald Giam在2024年提出的建議，以提供通用高階AI模型訪問許可權。SkillsFuture Premium AI Access Scheme為註冊選定課程的新加坡人提供六個月的免費工具，這是正確方向的一步。同樣，NTUC的補貼涵蓋21個AI工具。這些都是很好的開始，但它們不必要地受限於課程註冊和工會會員身份。至關重要的是，它們可能不會涵蓋AI代理——這是真實生產率差距將開放的層級。
為什麼這很重要？執行AI代理成本高昂。我們可能希望代理訪問的成本軌跡與網際網路頻寬或計算成本相同，但沒有必然的理由認為它一定會這樣。這是一個經驗性的問題。
Anthropic的執行長在1月表示，其80%的收入來自企業客戶，由按令牌付費模式的API呼叫產生。如果代理預設保持企業級別，那麼個人公民——求職者、自由職業者、退休者——將被排除在真正的生產率收益發生的層級之外。
有三個可能的方向。
其一，談判主權訪問權——與前沿AI提供商達成大規模許可協議，為所有公民提供量大從優的AI代理訪問權。
其二，如果代理訪問在市場中由僱主提供，則應使其普遍化，要求一定規模以上的公司為所有員工提供代理級別的AI，就像我們對CPF的要求一樣。
其三，如果前沿代理成本仍然過高，則確定最小的、可行的代理級別，併為其提供普遍資助。
政府是否會將高階AI訪問權作為普遍權利，而不是將其限制在課程完成或工會會員身份之後？
主席先生，我最近才瞭解到，甚至頂尖的兩三個前沿AI實驗室的AI工程師也擔心因為無法使用Claude Code而落後。我剛從中國回來，親身瞭解到在那裡根本無法使用Claude Code。Anthropic完全阻止來自中國大陸、香港和澳門的API呼叫。
如果甚至是開發前沿AI的工程師也渴望獲得彼此的工具，如果整個國家都可能被鎖定在外，那麼訪問權就不是一種便利。這是一種戰略能力。而我們國家面臨的問題是我們是否會確保這一點，還是永遠只能是價格接受者。
世界上也許有三到五家公司，其在定價、獲取和部署方面的決定將塑造每個經濟體的AI發展軌跡。當Anthropic或OpenAI決定對代理級別訪問的收費或其服務物件時，該決定對新加坡生產率邊界的影響與任何貿易協議一樣直接。
因此，我們應該以對待國家雙邊關係相同的戰略認真性來對待這類公司——已經跨越系統重要性閾值的前沿AI公司。不是因為它們是主權實體——它們並非如此，它們缺乏國家的永續性和合法性，仍然受母國法律約束——而是因為它們的決定對我們的經濟產生主權級別的後果，我們應該相應地與之互動。
這在實踐中意味著什麼？有四個方面。
首先，在主權層面談判獲取權。在前沿AI代理成本上升而非下降的可能未來中，新加坡應該尋求代理級別訪問的批次許可協議，就像我們談判能源供應一樣。這意味著要接受前沿AI訪問權可能成為國家支出中持續更高的專案，並有系統地採購，因為替代方案——公民因價格而無法獲取定義生產率的工具——更加糟糕。
其次，我們根據我們擁有的東西進行貿易。Nvidia執行長Jensen Huang將AI堆疊描述為五層蛋糕：能源、晶片、基礎設施、模型和應用。在我看來，我們在能源方面沒有規模優勢。我們沒有前沿模型能力。在應用層，除了我們能自己建立的知識聚集地外，幾乎沒有競爭優勢。我們將面臨與全球最高成本地區的競爭。
但新加坡在資料中心建設方面很有優勢。而且我們在水重複利用和綜合水管理方面處於世界領先地位，這對水資源短缺的東南亞地區的資料中心擴張形成了制約。如果我們將自己定位為這一地區首選的基礎設施夥伴，那就是真正的槓桿——這是我們可以用來交換訪問權、定價權和存在的籌碼。
因此，當Anthropic或OpenAI這樣的公司與我們接觸時，我們應該成為他們的首選區域雙邊夥伴，在區域內推出和擴大數據中心建設，以及提供這些資料中心運作所需的所有基礎設施。
第三，我們需要吸引真實的技術存在。我們應該尋求前沿AI公司在這裡建立開發辦公室——而不是主要建立銷售辦公室，這是2010年代FAANG公司的經歷。而且我更傾向於我們要有質量意識。大多數自稱為'AI公司'的公司並非前沿AI公司。我們需要針對前沿AI公司制定有針對性的戰略和合作。
第四，讓新加坡人進入這些實驗室。一旦你進入前沿人工智慧生態系統，就能更容易地在這些公司之間流動。我歡迎政府進行一些事實調查——聯絡已經在這些公司和職位上的本地和海外新加坡人，瞭解他們或同事是如何被聘用的，並將這些資訊傳播給我們的學生和技術研究人員。目前，根據傳聞，在美國，人工智慧研究人員的薪酬在50萬至100萬美元之間（不計股權）相當普遍。因此，很明顯，我們的利益所在就是想方設法讓更多新加坡人進入這個緊張的勞動力市場。我真正希望看到的是建立一套針對前沿人工智慧實驗室研究人員的技能框架。
主席，我最後一點是關於轉變的。讓我從一個人開始。在杭州，一位名叫周的質量保證主管在2022年末加入一家科技公司，月薪為人民幣25,000元，約新加坡元4,800元，負責審查人工智慧模型輸出的準確性和安全性。到了2025年，他的僱主決定可以由人工智慧模型來做他的工作。他們提供了一份薪資約降低40%的新崗位。他拒絕了。他們解僱了他。周申請仲裁併勝訴。公司隨後起訴並敗訴。公司上訴，再度在杭州中級人民法院敗訴。該判決書於今年4月28日發表，即國際勞動節前三天。
法院的理由值得我們關注。該公司主張，人工智慧已使周的角色過時——這是'客觀情況發生重大變化'，足以在中國《勞動合同法》下正當化解僱。法院不同意。法院認為，人工智慧的採用是一種刻意的商業戰略，而非不可預見的事件。選擇自動化的公司不能單方面將該決定的全部成本轉嫁給工人。該公司未能證明合同的履行實屬不可能，而減薪百分之四十的崗位調配也不是合理的替代方案。法院進一步指出，公司應優先對工人進行再培訓，並幫助他們向更高級別的職位過渡。
一項原則值得新加坡認真考慮,即故意的商業決策不應該將其全部成本轉嫁給工人。
如果一名新加坡人周明天在現有框架下遭裁員,他會贏嗎?我們現有的《管理過剩人力和負責任裁員三方諮詢委員會(TAMEM)》是諮詢性的,而非法定的。僱主可以合法地將一個職位自動化,並在不首先嚐試對工人轉崗或重新培訓的情況下,終止該工人的僱用。而公共財政——通過SkillsFuture和新加坡勞動力部——將承擔該工人轉變的成本。沒有人工智慧特定的通知期。沒有法定的轉崗優先義務。工人沒有訴權來對其解僱原因提出質疑。
資料表明我們正在進入這一問題開始產生影響的時期。人力部發布的2025年第四季度《勞動力市場報告》顯示,2025年的裁員人數約為14,490人,較前一年的約13,000人增加。PMET裁員發生率達到每千名常住員工10.1人,高於2015至2019年期間的衰退前常態8.0。裁員集中在金融服務、資訊通訊和專業服務這三個最容易受人工智慧影響的部門。在2025年,資訊通訊業就業出現了直接下滑。
政府在4月30日宣佈了三方就業委員會。我當然歡迎其意圖。但它沒有創造新的權力,沒有對僱主施加新的義務,也沒有給失業工人新的權利。政府打算如何讓這個委員會發揮作用?
工人所經歷的往往不是關於AI驅動重組的坦誠對話，而是績效改進計劃。在許多情況下，這個過程有些'虛演'的成分，旨在掩蓋一個預先確定的結果。我預見可能會給出這樣的誤導性理由，工人必須有權對此進行質疑。
我提議三個方向。首先，在AI驅動的職位消除前，強制規定90天過渡通知期。其次，優先重新部署的義務——在AI驅動的僱傭終止前進行重新培訓或調任。這些條款將放緩AI顛覆的速度，而速度決定了是否可能進行調整。第三，工人如果認為終止僱傭的理由具有誤導性，應能夠對其進行實質性的質疑，以確保這些AI重組保護措施真正有效。
議長先生,最後。芬蘭向人民提供了無條件現金收入。他們更幸福,壓力也更小——而且絕大多數人仍然走進就業辦公室申請工作。美國民調專家David Shor今年對美國人進行了民調:以三比一的比例,跨越各種政治立場,他們都選擇創造就業而非直接轉移支付。人們在被給予全民基本收入和就業之間選擇時,總是選擇就業。不是因為他們不理性,而是因為工作是你被需要的地方,而全民基本收入無法取代被需要的感覺。
所以,沒有失業增長——對的。但不僅僅是這樣:不能是收益被資本過度獲取、調整負擔落在勞動身上的增長。普遍獲得,所以知識不被財富所限。戰略性參與,所以我們在自己的未來中不是被動的價格接受者。以及一個裁員框架,其中決定自動化的公司先於工人承擔這一決定的成本。
我不認為敬畏和恐懼會消散。但在一個為工人而建設的國家裡，有更美好未來的希望。謝謝。
副議長先生：桑吉夫·蒂瓦瑞先生。
下午4時25分，桑吉夫·庫瑪·蒂瓦瑞先生（被提名議員）：副議長先生，在我開始之前，我向議廳兩側的各位工會同仁致以問候。感謝你們的支援。
副議長先生，作為工會人士，我們的角色不僅是支援變革，而是確保變革對我們的工人有利。從迄今為止的所有討論來看，毫無疑問，人工智慧帶來了新的機遇、新的工具、新的工作方式以及更好工作的潛力。
對於勞工運動而言，無失業增長的人工智慧轉型必須意味著三件事。第一，人工智慧增強工人的能力，而不是大規模替代他們。第二，人工智慧帶來的生產力收益被重新投資到人員培訓、新產業、更好的工資和實際增長中。第三，沒有任何工人被留下來獨自應對這一轉型。我將逐一論述這些問題。
確保生產力收益得到共享。副議長先生，如果我們不主動採取行動，人工智慧帶來的收益將不會自動得到共享。這些收益往往會集中在具有規模和能力來部署它的企業手中。轉型的商業案例令人信服，我支援政府宣佈的支援企業加快人工智慧採用的努力，以便企業能夠把握新的機遇。
然而，我呼籲尋求做大蛋糕的企業，要聚焦於工作重新設計和培訓，以便與工人一起推進。在這裡，我必須感謝許多支援這一呼籲的尊敬的議員。
全球有警告稱，人工智慧在未來幾年可能會大幅減少初級白領職位。在本地，星展銀行已宣佈計劃通過採用人工智慧，在各個市場將其合同員工和臨時員工共減少約4,000人。
雖然企業在最佳化員工隊伍規模和技能結構，但我們必須保持清醒的認識，即並非所有這樣的最佳化都能轉化為經濟增長，使我們的家庭興旺、孩子茁壯、老年人安享晚年。
因此，工人正在密切關注。他們想知道是否存在讓人工智慧的收益惠及工人而不僅僅是管理層和股東的途徑。許多工作場所才剛剛開始思考這樣的途徑需要什麼樣的形式，以及未能提供這樣途徑的隱患。
正如一些議員所提及的，最近中國法院一直在積極審查因人工智慧相關重組而解僱員工的案件，並選擇保護勞動權利，反對不公平的人工智慧相關裁員。在一個案件中，仲裁庭明確指出人工智慧替代不是解僱的合法理由。在另一個案件中，由於人工智慧接管工作導致的大幅職務和薪酬降低未被視為合理的重新分配建議。
我希望新加坡不會看到此類訴訟案件。因此，企業必須遵守人道轉型標準。當公司部署人工智慧導致職位被淘汰時，企業應該擁有轉型計劃、再就業機會、資助的再培訓和分階段的時間安排。
這些措施必須通過CTC或三方框架與工會合作進行，以確保共同管理。我們應該將其作為基準，而不是例外。僱主和員工之間的社會契約必須隨著技術而演進。
工會應該儘早參與公司的轉型計劃，以整合新技術、實施工作重新設計並協助工人適應新工作。在人工智慧未替代的職位中，人工智慧正在增加工作的速度、密度和複雜性，而非減少它。
但是，我要提醒大家，當人工智慧推動人類工作邊界而工作重新設計沒有充分同步進行時，人們應該擔心工作環境將變得不可持續，工作強度和壓力非常高，可能導致疲憊、疲勞和較差的心理社會健康。我們已經在許多當今工作場所看到這一情況，特別是對於專業、管理和行政人員（PME）而言，工作越來越多地基於結果，工作和個人生活之間的界限更加模糊。人工智慧存在進一步加速這一趨勢的風險。
人工智慧賦能的心理社會影響也值得進一步關注。我們基於時間的就業保護是為數字化程度較低的時代設計的，當時工作時間更固定，界限更明確。如今，這些新興工作模式表明我們在如何支援工人方面還有進一步演進的空間。
這是新加坡工會發揮關鍵作用的地方。通過集體協議、公司級別的接觸和多公司舉措，例如NTUC的CTC倡議中的蜜蜂皇后夥伴關係，工會確保生產力收益被轉化為更好的工作、更好的工資，最重要的是更好的工作條件，而不僅僅是更高的產出和股東回報。此外，當人工智慧被用於支援招聘決策和績效評估時，我們必須警惕無意中的偏見，並確保資訊的機密性保障。
通過轉型支援工人。即使收益得到共享，我們仍然面臨一個更棘手的問題：那些工作被徹底替代的工人會怎樣？
副議長先生，我們通常給出的標準回應是重新培訓、適應、繼續前行。這種建議假設工作者有時間、有財務緩衝和出錯的餘地來承擔這樣的風險。但並非所有工作者都有這種便利條件。我想要誠實相待，因為模糊的樂觀主義對於失業者來說是一種承受不起的奢侈，我們必須確保系統關注這一點。
對我們的中期職業和年長的PME來說尤其如此。他們是有抵押貸款和賬單要付的工作者，有子女仍在上學，通常還要照顧年邁的父母。他們不僅僅是在管理職業生涯。他們肩負著整個家庭的責任。
對他們來說，轉變是高風險的。一次失敗的轉變不僅僅是暫時的挫折。它可能意味著長期失業、收入損失以及對其家人的長期影響。這一點我們已經在人力部提供的資料中看到了，與2019年衰退前的規範相比，PMET裁員人數有所增加，反映出他們在經歷重組的行業中面臨更大的風險。
這就是我們必須進一步支援養家者及其家庭的地方，確保重新培訓導致真實的工作成果，轉變得到支援，通往好工作的路徑清晰，特別是對於那些進行中期職業轉變的人。
我們不是依靠福利，而是依靠工作福利。不是最低工資，而是對我們勞動力關鍵部門的漸進工資模式。在人工智慧時代，我們可以為我們的PME求職者尋求更好的支援。我們必須繼續密切監測那些申請了求職者支援的人，並幫助他們儘快重新找到下一個更好的工作。
副議長先生，在人工智慧採用中給予工作者真正的聲音，其中分享收益和支援轉變的前兩個支柱離不開第三個支柱。在人工智慧如何進入其工作場所的問題上給予工作者真正的聲音同樣重要。
在發達經濟體中，一個原則變得越來越清晰，工作者的聲音必須是技術如何重塑工作方式的一部分。在比利時等國家，工會和僱主已經在一起工作，圍繞工作時間之外的溝通、工作量和人員配置制定規範。
在新加坡，我們在三方模式中擁有堅實的基礎。但作為一名工會主義者，我想提出一個更廣泛的觀點。如果我們認真確保人工智慧被公平地使用，以及人工智慧驅動轉變帶來的收益能夠到達工作者手中，那麼我們必須歡迎工會代表在人工智慧時代最脆弱的PME。
PME不是一個單一的群體——航空航天領域的工程師、銀行業的金融分析師、科技階段的專案經理——他們的工作環境非常不同，經歷人工智慧轉變的影響也大不相同。工會能夠從基層塑造工作場所規範，以承認多樣性的有針對性方式來運作。因此，僱主應該考慮允許工會代表PME。
讓工作者聲音出現在桌子上的機制已經存在——它就是CTC。通過CTC，工會直接與管理層合作，制定轉變路線圖、重新設計工作和提升工作者技能，確保在公司轉變時沒有人被遺漏。
讓我舉一個例子。新巴士有限公司在全國運輸工人工會和CTC補助金的支援下，使用人工智慧全面改革了其公交車維護運營。該公司實施了人工智慧驅動的診斷系統用於預測性維護，而不是裁員，而是為超過50名工作者建立了一個新的診斷專家職業計劃。這樣的例子必須被放大，更多的僱主應該這樣做。
我們必須繼續利用這一點來支援工作者和企業在人工智慧轉變中的發展。最近，NTUC與亞馬遜網路服務（AWS）和華為等全球技術領導者合作，通過CTC生態系統為100,000名工作者和100家企業提供人工智慧技能培訓。
在個人層面，工會成員也可以獲得工會支援，支付人工智慧模型訂閱費的一半。這是除了政府提供的六個月訂閱之外的補充。我希望更多領先的跨國公司能夠與我們的工會合作，提供更多培訓，提升工作者的人工智慧技能，為我們的集體未來做出貢獻。
總之，副議長先生，我呼籲新加坡在這些領域向前發展。
首先，關於分享收益，我們必須確保當企業以人工智慧進行轉變時，收益通過公平工資、更好的工作條件和圍繞工作強度的更清晰規範與工作者分享，包括對工作時間之外溝通和響應的期望。這些必須根據我們的新加坡背景進行校準，但意圖明確。
其次，在支援轉變方面，我們必須加強我們如何通過人工智慧轉變支援工作者，確保人工智慧使企業能夠釋放新的增長，培訓導致真實的工作機會，重新培訓與工作重新設計結合在一起，工作者不會被單獨留下來應對這些變化。
第三，在給予工作者聲音方面，當人工智慧被引入工作場所時，工作者及其工會必須儘早參與。不僅僅是在招聘和其他就業決定上，而是在人工智慧如何改變工作範圍、工作流程和效能期望上。這意味著歡迎工會代表PME的能力，並擴大CTC等機制的規模，以便工作者的聲音嵌入在每個轉變旅程中。
沒有失業增長的人工智慧轉變不是一個口號。這是一個承諾。一個承諾，增長應該對每個人都有意義，而不僅僅是對經濟金字塔頂端的人。而是對護士、教師、物流工人、小業主、第一次進入勞動力的年輕人。他們不是技術進步故事中的腳註。他們是進步之所以重要的原因。這些不是相互競爭的優先事項；它們齊頭並進，我們正站在歷史上罕見的拐點之一，我們今天做出的選擇將為未來幾代人迴盪。
這就是三方合作伙伴必須履行、必須使其發生的地方，這也是我希望這個議院能幫助我們實現的。我強烈支援該議案。[掌聲。]
副議長先生：沙拉艾爾·塔哈先生。
下午4點38分 沙拉艾爾·塔哈先生（巴西立-樟宜）：謝謝您，副議長先生。副議長先生，我宣告我的利益，我從事航空航天和先進製造業工作，重點關注戰略、數字轉變和人工智慧轉變。
副議長先生，我今天站在這裡懷著深深的感激之心。我在一個非常普通的新加坡家庭中長大，父母都是工薪階層，但我獲得了在世界各地建造和改裝先進工廠的非凡機會。
從德國到英國，從北美到亞洲，我有幸從事工業4.0設施的建造和工作，配備了有史以來一些最精密機器的尖端工廠，這些機器生產的零件精度如此之高，只能在世界上為數不多的幾個地方製造，以及組裝了有史以來一些最先進的工程系統的設施。
副議長先生，經過這麼多年，一個教訓比所有其他教訓都更突出。決定成功的不是機器，也不是技術。是人。我見過擁有金錢能買到的最好技術的工廠因為錯位、因為不信任而陷入困境，因為工作者、工會、管理層和政府沒有朝同一個方向發展。我也見過更謙虛的設施超額完成預期，因為每個人都與共同目標相一致。
這就是為什麼這個辯論很重要，因為當我們談論轉變，特別是由人工智慧驅動的轉變時，我們不僅僅是在談論技術。我們在談論人、工作者、生計和尊嚴。
我想要承認我們的勞動運動、NTUC秘書長黃志明先生和楊晚齡女士以及議院成員李孟庭先生和沙克蒂亞迪·蘇帕特先生提出這項議案。他們的立場不僅僅是正確的。也是及時的。
它是及時的，因為它補充了總理勞倫斯·黃先生在其預算演講和五一勞動節集會上制定的方向，他在那裡談論了新加坡必須如何應對人工智慧和全球不確定性的現實，同時堅定地與我們的工作者站在一起。而且即使技術改變了我們的經濟，我們也不能把我們的人民落在身後。
我們的勞動運動以清晰和堅定的方式回應了這一點。這一轉變必須推動新加坡的下一個增長階段，但必須以所有人的公平和機會為基礎。我們必須裝備工作者和企業以抓住新的機會，使進步不僅僅是被創造，而是被分享。這是我們必須向下一代做出的承諾。
副議長先生，在不同國家工作過，我遇到了許多勞動運動，許多關注保護工作，保護今天存在的特定工作。通常為了保護今天的工作，他們不可避免地必須抵制變革，即使變革的浪潮是不可避免的。但我們在新加坡有什麼是不同的。
從我的企業經驗來看，NTUC、新加坡國家僱員聯合會和政府之間的三方合作伙伴關係是獨特的，獨特到我的許多海外同事都真誠地感到困惑，為什麼它能夠如此良好地運作，為所有人帶來真實的積極成果，因為我們的工會不僅僅保護工作，他們保護工作者。
我們的工會領導人,如電力和燃氣員工工會的薩瑪德兄弟、電子和電氣工業聯合工人會的法赫米兄弟、SATS工人工會的普博蘭和戈維登兄弟、法定委員會員工聯合工會的加布裡埃爾兄弟以及這裡的許多工會領導人,與工人站在一起,不僅是為了他們今天的處境,而是為了他們明天需要達到的位置。他們致力於保持工人的相關性、就業能力和準備好在經濟發展時抓住更好的機會。
副議長先生,這是我們永遠不應該掉以輕心的事情。請允許我圍繞三個關鍵理念來闡述我對這項議案的立場:轉變性力量、為所有人提供機會和無就業增長。
首先,關於人工智慧的轉變力量。人工智慧的影響可以在三個層面上理解:個人、企業和產業。在每個層面上,成功都取決於工人、企業和政府的合作程度。在個人層面上,人工智慧是一個力量倍增器。
它提高生產力、增強技能,使每個工人能夠做得更多、更好、更快。我們許多人已經通過Open AI、ChatGPT、Microsoft Copilot、Claude、Gemini和Canva等工具體驗到了這一點,但這種轉變不是偶然發生的。工人必須準備好學習、適應和不斷提升自己,企業必須投資培訓、重新設計工作並賦能其勞動力有效使用這些工具。
政府必須提供支援結構、強大的技能框架、可獲得的培訓途徑和廣泛的技術訪問權。這就是為什麼我對政府提供的幫助新加坡人採用人工智慧工具的支援感到鼓舞,這進一步由NTUC補貼其成員人工智慧訂閱的舉措加強。這很重要,因為人工智慧不能成為只有少數特權人士使用的工具。它必須對大眾保持可獲得,以便每個工人都有機會提高生產力、加強能力並有意義地參與新加坡的下一個增長階段。
在企業層面上,人工智慧可以實現更好的決策、更敏銳的運營和更高的效率。它將資料轉變為洞察力,將洞察力轉變為行動。但要實現這一點,公司需要正確的框架。工人需要適當的能力,政府必須提供正確的環境以負責任地擴大轉變。
在產業層面上,人工智慧創造了全新的價值。它重新塑造商業模式。它改變競爭並解鎖新的增長。在新加坡這樣的緊張勞動力市場中,這種企業和產業轉變必須與深入的業務流程重新設計和有意義的工作重新設計齊頭並進,正如這裡的一些議員所提到的。
根據我在該行業的親身經歷,採用人工智慧的決定並不僅僅是關於稅收激勵,正如議員Andre Low先生所指出的那樣。真正的驅動力是我們如何提升和重新培訓我們的工人,使他們能夠承擔需求不斷增長的高附加值工作,以及不這樣做的機會成本。
稅收激勵幫助公司投資必要的人工智慧工具和基礎設施。但同樣重要的是支援工人度過這一轉變的方案,無論是SkillsFuture勞動力發展補助金、NTUC CTC補助金、工會培訓援助計劃和新加坡勞動力部的職業轉換計劃等計劃,這些舉措支援工作重新設計、培訓,甚至在工人進行提升和重新培訓期間提供工資支援。
通常,採用人工智慧並不是技術和工人之間的二元選擇。這是關於重新設計業務流程,以便企業能夠通過更有能力、更熟練和更高效的勞動力做更多事情來應對世界的挑戰。
這些由人工智慧解鎖的價值必須被分享——分享以便工人得到提升、企業變得更強大、社會一起向前發展。但最終,這是我們必須繼續維持的工人、企業和政府之間的社會契約——一個以公平性、包容性和共同進步為基礎的契約。
綜合考慮,人工智慧不僅僅是任何其他技術工具。它是一個全系統的轉變。這就是為什麼三方合作在人工智慧時代更加重要。我很高興聽到勞工運動的立場。
第二,關於為所有人提供機會。議長先生,這一轉變將創造經濟增長。但這也是一個充滿真實不確定性的時期。我們必須承認這些挑戰。我們的應屆畢業生正在感受到這一點。許多人為了確保永久職位而苦惱。多個實習正在成為常態。我們在什麼時候會詢問這是否成為真正就業的替代品?
當人工智慧重新塑造從開發者到律師的職業時,我們如何確保我們的年輕人仍有有意義的入口點?如果公司開始質疑入門級角色,那麼我們也必須問,誰將培訓我們下一代的勞動力?
我們的職業中期工人感受到這一點更為深刻。他們有家庭和責任,他們擔心工作轉變可能使他們的經驗不那麼相關。我們的藍領工人——我們的技術人員、操作員和司機——對自動化和他們工作的未來提出了尖銳的問題。
這些恐懼是真實的。如果不解決,它們可能會分裂我們的社會。這就是為什麼這種為所有人提供的機會不能僅僅留給市場力量,這也是議員Poh Li San提出的一點。它要求公司,特別是中層管理人員,給應屆畢業生一個真正的機會。它要求企業重新設計工作並真誠地投資提升和重新培訓。它要求我們所有人一起工作,以便每個新加坡人都能在這個新經濟中找到自己的位置。只有這樣,我們才能說這不僅僅是增長,而是為所有人提供的機會。
最後,關於無就業增長。在許多經濟體中,無就業增長意味著沒有就業的增長。但在新加坡,我們的情況不同。我們已經接近充分就業。所以,這不僅僅是關於創造更多的工作。這是關於確保我們的人民能夠承擔創造的工作。
因為增長將會到來,新的角色將會出現。但如果我們的工人還沒有準備好,如果技能沒有跟上步伐,我們就會冒一種不同的無就業增長的風險——不是缺乏工作,而是工作與我們勞動力擁有的技能之間的不匹配。避免這種情況意味著關注能力建設,而不僅僅是人員數量。
企業必須改變工作性質,而不是消除它們。工人必須繼續學習。政府必須通過強大的系統和途徑支援希望。
當我們做好這一點時,增長不會讓人們掉隊。它會提升他們。副議長先生,請用馬來語。
(用馬來語):【請參閱方言演講。】議長先生,關於人工智慧的討論今天已經不再是離我們生活很遙遠的事情。在新加坡和世界各地,一件事是明確的——人工智慧帶來了希望,但也帶來了關切。
在我們的馬來/穆斯林社群中,這些關切是真實的。年輕人擔心就業的未來以及是否仍然存在機會。那些職業中期的人對自己的經驗和技能是否仍然相關感到焦慮。藍領工人——包括司機、技術人員和普通工人——想知道他們的工作是否會被自動化取代。
我們必須承認這些關切。但與此同時,我們不能僅把人工智慧視為威脅。我們必須將其視為一個機會——一個一起進步的機會。
請允許我涉及人工智慧轉變中的三個重要點。
首先,關於建立技能以成為這一轉變力量的一部分。人工智慧只有在我們知道如何使用它時才會成為力量倍增器。這就是為什麼我們必須準備好不斷提升自己——重新培訓和提升技能。
這裡重要的是我們利用可用的支援——無論是通過新加坡勞動力(WSG)計劃、NTUC還是社群舉措,如M3+。例如,在M3+ Pasir Ris–Changi,我們運營各種計劃來幫助社群提高他們的技能和就業機會。這些包括女性工作計劃,幫助女性重返勞動力市場,以及在Pasir Ris舉辦的職業市場,以開放獲取職業機會和就業網路。
這些努力沒有停止在那裡。通過Pasir Ris–Changi的HashTech計劃,我們開始向我們的孩子介紹人工智慧、機器人和自主系統——包括通過機器人大戰等活動。這不僅僅是一個活動。這是一個從小開始建立信心、接觸和未來就緒技能的努力。
關於技能,對人工智慧的恐懼不應該阻止我們學習如何明智地使用它。人工智慧可能削弱對馬來語言和文化以及其他技能的深入理解的關切是可以理解的——如果在不理解基礎知識和其侷限性的情況下使用它。
像任何其他工具一樣,重要的是我們如何使用它。掌握和欣賞馬來語言在理解其美麗、價值和意義中仍然是關鍵。
但是,如果使用得當,人工智慧也可以幫助儲存我們的遺產。人工智慧可以協助數字化舊Jawi手稿,生成受傳統馬來圖案啟發的蠟染設計,並促進馬來語的學習和翻譯。甚至經典的馬來電影也可以為未來幾代人保留。
技術不應侵蝕我們的身份認同。如果明智地使用，人工智慧可以幫助保護語言、加強文化，並自信地將馬來傳統傳承到未來。
請允許我也分享一首班頓詩。
金色香蕉送到海，
一個在箱頂成熟，
如果人工智慧被非常明智地使用，
文化代代相傳，深植心中。
(英文)：副議長先生，從根本上說，這不僅僅是一次經濟轉型。這是對我們社會契約的考驗，這份契約必須在新時代重新確立，在這個新時代中，企業的承諾不僅僅是利潤，更是人民；工人的承諾不僅僅是工作，更是終身成長；政府繼續與兩者並肩同行，確保沒有人被落下。
如果我們能夠做到這一點，如果我們能夠以信任、目標和共同責任一起向前邁進，這項轉型就不會分裂我們。它將使我們更加強大，因為當工人、企業和政府一起行動時，我們不僅僅是適應變化。我們塑造變化，我們從中受益，我們確保每一個新加坡人都一起向前邁進。副議長先生，我支援這項動議。[掌聲。]
副議長先生：副教授 Terence Ho。
下午4時54分 副教授 Terence Ho（被提名議員）：副議長先生，我起身支援這項及時的動議。我首先想宣告我的利益關係——我是新加坡社會科學大學成人學習學院的執行董事。
眾所周知，人工智慧是一種變革性技術。我們都認識到，它將深刻改變整個經濟中的商業模式，進而改變工作的內容和性質。
作為一個小而開放、技術先進的國家，新加坡必須努力走在新技術的前沿，特別是像人工智慧這樣重要的技術。除了對我們的經濟至關重要外，人工智慧在應對新加坡人口和社會挑戰方面也具有相當的潛力。但無論我們擁抱還是畏懼人工智慧，都無法逃避它對我們公司和員工隊伍的影響。
我將在演講中提出四點。
首先，無就業增長不是新加坡的選擇，因為優質就業對包容性增長以及實現公平、充滿活力的社會至關重要。
新加坡的社會契約基於通過工作實現自給自足。這意味著通過就業和收入為自己和家人提供生活保障。多年前曾在本院表述過，'工作是最好的福利，充分就業是新加坡工人的最好保護。'
在這份社會契約中，政府的角色是培育有利於投資和創造就業的親企業商業環境。任何願意辛勤工作的新加坡人都將通過中央公積金儲蓄有足夠的資金來滿足住房、醫療保健和退休需求。在實踐中，政府以住房補助、中央公積金增值和醫療保健補貼的形式提供重大支援，以支援房屋所有權、退休充分性和醫療保健保障。
新加坡的社會經濟模式隨著時間推移而演變。我們現在通過社會保險實現更廣泛的風險共擔，補充個人儲蓄以滿足醫療保健和長期護理需求。存在更多結構性或永久性的社會支援，包括工作收入補充和銀髮支援。政府認識到經濟和就業中斷的更大風險，已通過SkillsFuture求職者支援為非自願失業者引入收入救濟，這已由議院議員討論過。
但就業和收入仍然是新加坡社會經濟模式和社會契約的核心。人工智慧的最近進展對這一模式構成了挑戰。全球享譽的人工智慧先驅和行業領袖曾警告人工智慧可能導致大規模失業的可能性。所謂"白領血洗"或"就業末日"的預測已激起公眾恐懼，儘管其他評論人士聲稱這些恐懼過度了。
過去的變革性技術確實消除了某些工作，但它們也創造了新工作，因此我們不會都沒有工作或過著無限閒暇的生活。
生成式人工智慧引起了特別的關注，因為它可以接管與人類技能和創意相關的工作和任務，包括編碼和資料分析等認知任務以及寫作和設計等創意任務。這些任務需要通過多年教育和培訓建立的技能，因此獲得很好的報酬。它們支撐了許多新加坡人渴望的好工作。
雖然我們不應該低估人工智慧的破壞，但仍有時間適應和應對。這是因為工作流失或失業的程度取決於技術擴散的速度，這在歷史上遵循了S曲線。
我在3月提交了一份議會質詢，詢問人力部是否發現了任何人工智慧促使新畢業生招聘放緩的跡象。確實，議院的許多議員都指出了這一關切。我當時收到的回應，昨天也得到了重申，是年輕學位持有者的就業率保持基本穩定。
這可能是因為人工智慧的採用對許多本地公司仍處於早期階段。許多公司需要時間從將人工智慧視為僅僅的專案或生產力增強工具轉變為圍繞人工智慧從根本上重新組織工作流程。然而，隨著人工智慧採用步伐加快，對公司生產力和利潤的好處將增長，但對工作的影響也會增加。
許多議院議員也討論過一個相關的關切，即在人工智慧時代，利潤將越來越多地流向科技公司和這些公司的股東，而工人獲得的利潤將減少，因為技能變得商品化。
雖然我們必須考慮對更多社會支援和新的再分配渠道的需要以保持我們社會的包容性，我們也必須繼續為公民賦予通過良好工作和收入自給自足的能力。
我們中的一些人可能還記得Tharman總統在2015年聖加倫研討會的一次採訪中所解釋的內容。他在描述新加坡的方法，我引用道："這是關於保持活力的文化，其中我為擁有自己的房子感到驕傲，通過我的工作賺取我自己的成功。我為撫養我的家人感到驕傲。保持這種文化的延續是使社會充滿活力的原因。"
如果我們看世界各地經濟如何發展，很明顯成功的關鍵是由教育和創造就業機會推動的強大中產階級的出現。許多擁有豐富自然資源的國家表現不佳，因為他們的重點一直是資源開採，使少數人受益，而不是教育和技能發展，這使很多人受益。也很明顯，新加坡之所以成功，正是因為人民是我們唯一的資源。新加坡的經濟發展是包容性增長的故事——這是我們必須堅持的道路。
這引導我進入第二點，我們必須支援工人發展深層領域知識、學習敏捷性和職業韌性。為使工人適應人工智慧時代，人工智慧工具培訓當然很重要。對人工智慧的熟悉和對不同人工智慧工具的優勢和侷限性的理解來自頻繁使用、除錯和實驗。但是，正如議院許多議員也意識到的那樣，這只是部分答案。畢竟，人工智慧工具應該隨著時間的推移變得更直觀和易於使用。
人們為工作帶來的真正價值在於對領域或主題的深入理解，這意味著我們所有人仍然必須投入艱苦的努力來學習，並通過在學習過程中創造建設性摩擦來避免"認知解除安裝"，無論是在學校還是在工作場所。人工智慧可以提供腳手架並可以幫助個性化學習，但它不能成為思考的替代品。
許多人確定的關鍵技能是學習如何學習，對變化具有適應性和韌性。這意味著走出我們的舒適區，不斷挑戰自己，適應不同的任務和工作環境，以及在多樣化團隊中工作。
這是每個人的責任：對自己的學習和職業發展負起責任，並在可用的情況下充分利用公共資金和資源。
僱主也負有責任。由人力部(MOM)開發的 Singapore Opportunity Index 突顯了僱主如何為員工創造機會，例如通過招聘實踐、職業發展路徑和工作設計。通過識別具有支援職業成長良好記錄的僱主，希望能夠鼓勵這些僱主投資於員工，幫助他們規劃職業生涯。
第三點我想提出的是，我們不僅要為工作提升工人的技能，還要為工人改善工作。正如我在MOM撥款委員會辯論的演講中所提到的，在醫療保健和技能行業等領域，對工人的需求將繼續保持強勁，這是由老齡化人口以及這些職位相對抗AI取代的韌性所驅動的。
然而，這些工作很難吸引新加坡人，因為相比白領辦公室工作，它們被認為聲望較低或回報較少，或許不夠舒適。但風險在於，當新加坡人努力尋找符合他們期望的工作之際，新加坡卻日益依賴外國人來承擔必要工作。
目前，本地勞動力中超過40%擁有大學學位。因此，傳統上被認為不需要學位的工作和職業必須重新設計，以適應包括畢業生在內的各種勞動力。
薪酬只是問題的一部分。工作必須重新設計，以開發工人的"頭腦、心靈和雙手"技能，使其對工人更具吸引力，對人工智慧顛覆具有更強的韌性。這可能涉及承擔更大的專業責任、增加工作的認知、分析和創新內容，以及在工作中創造更多人際互動的機會。人工智慧工具實際上可以通過增強人類專業知識來支援職業和半熟練工作的升級。通過使這整個範圍的工作對新加坡人更具吸引力，我們可以避免結構性過度資格或就業不足。企業可以通過以技能而非資質為中心來進行工作設計。
我的第四點是，新加坡應該建立專業知識，成為技能優先和以人為中心工作重新設計的全球參考點。技能優先的方法意味著工人不會因其正式資格而被限制在特定的工作或職業。僱主認識到，無論工人的起點如何，他們都可以通過提升技能來滿足工作要求。同樣，工作和職業的範圍可以擴充套件，以更充分地利用工人的技能。
人工智慧已經在對工作進行分割——分解為多項任務，其中一些分配給人工智慧，另一些分配給人類工作者。重要的是，工作重新設計應該以人為中心，這樣人類工作者仍然可以做出有價值的貢獻，由人工智慧和技術增強，而不是讓流程完全自動化，人類參與最少。
隨著智慧代理AI現已能夠執行跨越多個工作角色的流程，僅重新設計特定工作內的任務已經不夠了。相反，組織必須著眼於重新設計端到端的工作流程。這是一種必須嵌入組織內部的能力，因為AI不斷重塑工作的本質。新加坡有機會在技能優先就業實踐和以人為中心的工作重新設計方面引領步伐，這將同時造福我們的企業和勞動力，對我們的社會契約也至關重要。
成人學習學院的技能優先實踐中心最近推出了一系列技能優先工作論文和配套的圓桌討論。這些論文和討論吸引了來自世界各地的專家、政策制定者和行業從業者的廣泛國際關注。
人工智慧對工作和學習的影響是所有國家和社會都在努力應對的一個問題，沒有人擁有所有答案。這是一個富有前景的研究領域，因為企業和社會都在尋求適應和轉變。我們處於一個複雜的運營環境中，沒有久經考驗的方案可以依靠。正如我最近與一位企業領導人交談時所說的那樣：我們在飛行中學習和適應。
在人工智慧時代，學習和工作重設計都必須是迭代的、實驗性的。我們需要新加坡和世界各地最聰慧的人才專注於此。正如新加坡在量子計算等新技術領域擁有世界領先的研究中心一樣，我們應該在成人學習和以人為中心的工作重設計方面積累專業知識和經驗。
當今，高等教育機構中遍佈著卓越中心。例如，理工學院和ITE屬於職場學習的全國卓越中心。新加坡管理大學已建立了韌性勞動力研究所，而新加坡科技學院推出了技能評估和驗證倡議。在新加坡社會科學大學，成人學習研究所設有成人學習協作中心和技能優先實踐中心。憑藉從研究到轉化的全方位專業知識，成人學習研究所可成為成人學習和就業能力的全國中心。
我們可以共同將新加坡建設為成人學習、職業健康、人機互補性和以技能為先的實踐等領域的思想領導者和實驗室。作為一個融入國際尖端研究和實踐網路的全球創新者，新加坡可以幫助以促進經濟增長和人類繁榮發展的方式塑造未來工作。
副議長先生，擺在我們面前的動議有力地表達了新加坡三方合作伙伴對包容性發展的承諾，這種發展造福工人和企業。這很重要，因為我們不能再理所當然地認為未來的經濟增長必然伴隨充分就業和收入上升。除了適應人工智慧，我們的最終目標必須是使每位工人能夠成長、做出貢獻，並在工作中找到意義。
國家人工智慧委員會和三方就業委員會的成立體現了這一承諾。這是一項舉國上下的努力，每個人都有一份力量可以貢獻。
最近，一位美國記者接觸我，想了解新加坡如何應對人工智慧及其對就業的影響。她指出，許多大型美國公司面臨大幅削減員工的壓力，以此來提高利潤。確實，股東和投資者期望管理層用人工智慧替代員工。當我向她分享新加坡的優先事項是培訓員工與人工智慧一起工作時，她用她的話把這種對比描述為「鮮明」和「鼓舞人心」。
藉助我們獨特的三方合作以及所有夥伴的承諾，我相信企業和員工能夠以堅定的決心和信心迎接即將到來的人工智慧轉型。
副議長先生：楊亞歷克斯先生。
下午5時09分 Alex Yeo先生（Potong Pasir）：副議長先生，人工智慧可能令人害怕。我想起工作中第一次收到年輕同事的法律意見書初稿，其中某些內容是由人工智慧生成的，當時我心中警鐘大作。「這些內容可靠嗎？我會因為提交這份檔案而被法院問責嗎？」儘管這些內容已經過核實，並與經過深思熟慮的「人類」生成的法律分析結合在一起，但我仍然感到焦慮和擔憂。
這個事件讓我回想起我初入法律界時的情景。我會在辦公桌上收到來自某位高階合夥人的書面備忘錄，其中包含關於某些事項的指示。有一次，我問他是否可以用電子郵件會更容易，但我被告知電子郵件不可靠，而有了書面備忘錄，他才能確信我會收到資訊。我們都知道不能和老闆爭論，但最終這些書面備忘錄確實轉變成了電子郵件，也許是因為意識到電子郵件指示即使在辦公時間之外也能可靠地傳達給我。
正如歷史上每次工業和技術變革一樣，無論是蒸汽機、電力、個人電腦帶來的數字化、網際網路帶來的連線，還是現在的人工智慧，變革和轉變總是伴隨著焦慮和對未知的恐懼。人類天生對我們無法控制的事物抱有不信任——也許這樣甚至是對的。
在人工智慧方面，我們在新加坡尤為深切地感受到這一點。議院的許多議員都談到了這些顧慮，以及來自各行各業新加坡人的憂慮。俗語有云，「唯一不變的就是變化」。對未知的恐懼是一種自然反應，但我們應該將其作為一種激勵力量來重新思考舊的方式、學習新的方式，進而抓住新的機遇並實現增長。
因此，這項動議是及時的。它承認，利用人工智慧促進增長是一把雙刃劍。一方面，人工智慧可以成為新加坡經濟發展下一階段的驅動力；另一方面，如果其發展和部署不加約束，可能導致社會問題，如就業流失和不平等加劇。
正如總理在其預算演講中指出的那樣，人工智慧只是一個"工具"。我們如何利用它以及如何管理其部署將塑造我們的經濟、我們的工作、我們的生活。因此，正如動議所述，我們對AI驅動增長的方法必須以公平、韌性和全民機會為基礎。這項動議引起共鳴，因為它關乎將人民置於新加坡AI驅動增長方法的中心。增長必須具有包容性，造福我們的人民。我們不能不惜一切代價追求無就業增長。
先生，人們對人工智慧對工作場所和勞動力的破壞性影響的認識程度極高。我們各行業的工人和專業人士、經理及行政人員(PMEs)的憂慮和關切不是假設性的。這些都是真實存在的，甚至是可以量化的。新工會(NTUC)最近進行的一項調查發現，我們超過半數的PMEs感到迫切需要提升技能以保持相關性。近三分之一的人對被替代感到焦慮。其他研究表明，半數新加坡人擔心他們的角色可能被自動化，許多人關切人工智慧將更多地惠及公司利潤而非普通工人。
我們初次進入勞動力市場的年輕畢業生面臨著嚴峻的現實。隨著人工智慧對日常工作的自動化，僱主們提高了用人標準，從一開始就要求具備更高階的批判性思維和人工智慧協作能力。如果任其發展，人工智慧的顛覆可能導致不公平的結果。那些能夠搶佔先機或成功過渡的組織和個人將獲得指數級增長的收益，而那些未能做到的將被落在後面。
多年來，我們一直主動預判像這樣的顛覆性變化。通過SkillsFuture等專案，我們向新加坡人投資並灌輸終身學習的價值觀，以及定期提升技能的必要性。這為我們現在大力推進AI採納奠定了良好基礎，如國家AI影響計劃（National AI Impact Programme）旨在支援一萬家企業，幫助十萬名工作者提升AI能力。
然而，普通民眾的AI能力和素養同樣重要。
對此，我想提出兩點。
首先，你們會想起我演講開始時講的個人逸事。無論是過去的計算機、智慧手機，還是今天的人工智慧，熟練使用任何工具都關乎建立信心。
林文興先生是前NTUC秘書長，他與我分享了他在上世紀80年代參與政府計算機化工作的經歷。這在當時是戰略性決定，但工人們害怕計算機。因此，政府設計了計算機素養課程，由NTUC推出，使用早期蘋果計算機。熟悉鍵盤的使用，以及對於那些記得的人來說，玩PacMan這樣的遊戲。工人們逐漸克服了恐懼，擁抱了熟悉感。重要的是，關鍵資訊必須是，這是一個能幫助你更好、更快地完成工作的工具。現在，這個新工具是人工智慧。
雖然提高我們工人的技能以利用與各自工作場所相關的人工智慧工具至關重要，但人工智慧素養應該是一項國家級的努力，應在我們的學校、社群俱樂部甚至積極老齡化中心提供。
這個想法是將人工智慧素養融入生活的一部分，無論是使用人工智慧工具建立一個簡單的電子賀卡，該工具可以建立動畫圖形，或使用機器人建立一個十億美元的公司。只有這樣，我們才能設想整個人民在具有人工智慧驅動增長的經濟中共同進步。
如果擁抱人工智慧是一項國家戰略舉措，那麼我們應該為所有新加坡人推出一個國家人工智慧識字和素養計劃。
其次，支援工人成為人工智慧流利使用者應該不僅僅是為他們提供人工智慧工具和使用工具的方法。我們還必須為工人提供時間和能力，以學習如何在他們的工作場所應用人工智慧工具。
讓年輕畢業生找到工作，並確保年長工人和PME進入新的角色是迫切需要的，但如果人工智慧要改變我們的經濟，我們必須確保我們的勞動力通過更深層次的學習有效地學習，同時說服我們的企業和組織建立允許我們的工人測試他們的新技能、從錯誤中學習和改進的工作場所和系統。學習必須與建設能力相結合。
在最廣泛和最普遍的工具使用中，人工智慧可以在幾秒鐘內生成內容、總結和回答問題。我們需要一支不僅能夠使用人工智慧工具獲得這些成果，而且能夠與人工智慧作為協作夥伴合作，同時應用判斷力、理由、創造力和背景來推動高影響力價值的勞動力——人類元素。
何銘華教授在最近一篇倡導更深層次學習的《商業時報》文章中舉了美光新加坡的例子，該公司的內部人工智慧技能提升計劃不僅停留在意識階段，還讓員工使用人工智慧工具更快地提取見解、更好地分析風險、自動化日常任務、更有效地規劃專案和改進決策制定。
我們如何說服更多的企業和組織投入時間和寶貴的資源與我們一起實現這些目標？提高我們的勞動力技能，確保他們擁有必要的憑證和知識，可能會讓他們找到工作，但這不能保證企業和組織會以能夠建立和最佳化他們的人工智慧能力的方式僱用和培訓工人。
本政府一直刻意地將我們的政策重點放在長期進步而不是短期收益上。因此，我相信由總理領導的人工智慧國家委員會將以相同的基礎制定計劃和倡議。
事實上，在我們的制度中，我們已經具備了政府、工會和僱主密切合作的能力，以實現長期目標，確保我們的勞動力配備了知識、技能、深刻理解和實際應用成果。
我們的三方合作模式，其中所有三個合作伙伴經過幾十年建立了信任、相互尊重和平等夥伴關係，在人工智慧採用的這個新階段攜手合作，將幫助我們確保我們的工人、企業和經濟能夠抓住新機遇並共同進步。
在我結束之前，我想支援NTUC秘書長黃志明先生關於建立針對新加坡市場的市場情報和前瞻系統的提議。我們可以理解，從三方合作伙伴的資訊、資料點、分析和研究中汲取見解，以感知早期訊號、協調應對和在必要時提供主動早期干預，這將有多麼有用。
話雖如此，雖然支援我們的工人重新培訓並轉向不同的角色以防止失業很重要，但我們也應該探索使用相同系統中的資訊來確定政府如何激勵人工智慧初創企業為我們的勞動力創造新的就業機會的可能性。例如，據報道，中國的人工智慧電影製作初創企業得到了激勵和財政支援。據報道，他們開創了微劇或豎屏劇這種新的娛樂格式，在全球掀起了波瀾。
先生，如果我們要以人工智慧作為新加坡經濟發展的下一個階段，我們必須進行轉型，不僅要改造我們的經濟，還要改善每一位工作者和每一位新加坡人的生活。副議長先生，我支援本動議。
副議長：Patrick Tay先生。
下午5時20分 Patrick Tay Teck Guan先生（先鋒區）：副議長先生，我起身支援本動議。我想感謝本議院中的NTUC兄弟姐妹為支援本動議而到場。
像PME（專業人員、經理和行政人員）這樣的知識工作者極易受人工智慧影響，不同於之前主要影響基層員工的自動化浪潮。我們許多中等收入的PME屬於被夾在中間或保護不足的群體。他們被期望表現得像高管一樣，但保護程度不如基層員工，同時還要面對來自外籍PME的激烈競爭。處於職業中期的PME被夾在年輕和年長的受贍養人之間，承受不起失業。然而，當他們因收入和年齡較高而被裁員時，往往需要花費更長時間才能找到新工作，通常還要付出降薪的代價。
我們認為PME具有特權地位和適應能力，他們擁有資源，能夠主動承擔責任以提升技能，也能夠應對挫折。這一假設已不再成立。人工智慧將在所有PME部門和各個級別中既增強又破壞工作任務。我們必須認識到人工智慧是一種變革性技術，它有潛力創造新機遇和共同繁榮，但也有潛力通過將財富和權力集中在頂端來加劇不平等，特別是如果在採用人工智慧的競爭中忽視了對廣大中產階層的保護措施。
為此，我主張新加坡對人工智慧的政策必須以人為中心，而由人工智慧推動的增長必須以工作者為中心。這意味著我們致力於我所稱的'3E'。'3E'是公平增長、加強的保護和參與的勞動隊伍。
第一個'E'，公平增長。當我們談論人工智慧驅動的無就業增長時，這並不必然意味著大規模失業。然而，這可能意味著人工智慧帶來的收益可能不會流向廣大中產階層。今年早些時候，經濟發展局（EDB）宣佈預期創造的就業崗位數下降至15,700，這是至少20年來最低的數字，儘管所吸引的投資比前一年更多。這表明雖然我們可能看不到無就業增長，但我們很可能會看到創造崗位較少的增長。求職者可能會為更少的空缺崗位進行競爭，就業不足可能會上升，工資可能會停滯。與以前相比，年輕畢業生可能會面臨更大的挑戰來獲得全職就業。
在全球範圍內，我們已經看到一批公司，特別是科技部門的公司，宣佈了大規模的裁員，原因是人工智慧。
公平增長意味著AI驅動的增長所得將與工人共享，形式包括更高的工資、福利和工作前景。作為開始，我們需要提高對何為公平和負責任裁員的標準，例如要求提前裁員通知、支援工會根據行業規範談判裁員福利，以及設計AI補助激勵措施，要求僱主證明他們為有意義地重新部署那些工作被AI接管的工人而做出的努力。
當新加坡努力吸引全球頂尖AI人才來到我們的國家時，我們也必須確保這建立並加強我們的新加坡核心。我要求政府通過諸如能力轉移計劃等專案鼓勵互惠，以便知識和專業知識流向我們的本地中小企業。同時，政府也可以通過諸如海外市場沉浸計劃等專案，將本地AI人才送往海外頂尖AI公司進行培訓實習，從而投資於本地AI人才。引進全球專業知識與開發我們自己的全球人才管道是同一枚硬幣的兩個方面。兩者都深化了我們本地勞動力的能力。
第二個"E"，增強保護。在向AI的過渡中，不可避免地，一些工人會受到影響。我們不能讓他們自生自滅。他們需要職業指導、財務支援和機會來更強地反彈。我感謝政府推出求職者支援計劃，讓非自願失業的工人受益於長達六個月的過渡性支援，同時他們接受培訓或尋找下一份工作。
求職者支援計劃的收入門檻目前設定為5000美元（不包括中央公積金），這隻能覆蓋少於20%的本地中小企業。我希望政府考慮將此門檻提高至本地中小企業的總中位數收入，目前為2025年的8400美元，或考慮一些其他合適的等效方案來滿足受影響中小企業的需求。同樣地，政府也可以考慮允許擁有HDB住房貸款的人延期還款長達六個月，特別是當他們無法支付抵押貸款時，以緩解他們的即時現金流需求。
新加坡採取了經過深思熟慮、框架化的人工智慧治理方法。我們沒有落後。我們在謹慎推進。通過資訊媒體發展局的人工智慧治理框架、網路安全域性為保護代理人工智慧系統的補充意見和自願測試工具包，我們鼓勵了創新和負責任的採用。
但隨著人工智慧從協助決策轉變為制定決策，包括在招聘、晉升和重組中，我們必須跟上步伐。其他發達經濟體已經率先採取行動，明確將與就業相關的人工智慧列為'高風險'。
這對工作者意義重大。招聘AI是否會自動將殘疾應聘者排名較低？如果人力資源部門在重組期間依賴AI工具來篩選誰留誰走，當結果看似不公平時，工作者有什麼救濟途徑？在這種情況下誰應當承擔責任——人力資源團隊、採購該AI工具的僱主，還是開發該工具的開發者？
這些不是假設性的關切。新加坡科技設計大學（SUTD）與GovTech聯合研究發現，大規模學習模型可以從愛好和志願服務等其他資料點中可靠地推斷出個人特徵，如候選人的性別，即使簡歷已進行匿名處理。
我們很幸運通過了《工作場所公平法案》。其原則——即就業決定必須公平並基於勞績，每位工作者都有權獲得公平的好工作機會——無論這些決定是由人類還是演算法做出，都應該適用。
但是，雖然這些原則是技術中立的，我們目前的反歧視措施還沒有明確解決如何將它們應用於人工智慧做出的決定。隨著人工智慧的應用加速，僱主需要明確什麼是負責任的使用，工作者需要確保現有保護措施在進入人工智慧支援的工作場所時仍然適用。我們應該縮小這一差距，不是通過扼殺創新的法規，而是通過清晰、實際的三方指南來確保公正和公平的過渡。新加坡的優勢是我們的三方傳統，我們應該加以利用。
為此，我請求政府考慮以下幾點。首先，採用就業相關人工智慧的僱主應該被指導進行與影響程度相適應的風險評估，並確保有意義的人類監督。其次，人力資源專業人士應得到培訓和自我評估工具的支援，以負責任地使用人工智慧並識別偏見。參與資料治理和網路安全的官員也可以在這個不斷發展的領域更新他們的技能和知識，因為越來越多的公司採用與企業資料整合的人工智慧。第三，應該給予工作者透明度。他們應該知道在影響他們的決定中人工智慧在哪裡被使用，什麼防護措施已到位，以及現有的救濟途徑如何適用。第四，我們應該從源頭著手，與人工智慧供應商和開發者合作，以確保底層軟體在到達我們的工作場所之前符合公平和透明的基本原則。國際勞動組織已經開始了這一旅程，新加坡憑藉我們的三方傳統，可以成為這個領域的先行者。
第三個也是最後一個"E"，即敬業的勞動力。工作者掌握人工智慧並弄清楚如何將人工智慧融入他們的工作流程，而不是反過來。最終，一個組織不能只靠人工智慧執行，必須有工作者參與。簡單地從上往下強制推行人工智慧不會產生效果，甚至可能產生牴觸和沉沒成本。工作者是終端使用者，也是他們自己工作流程中的專家。他們知道人工智慧在哪裡增加價值，在哪裡不足。如果你想讓人工智慧發揮作用，你必須詢問做這項工作的人。
我們已經有了一個經過驗證的機制來做到這一點——我們全國工會大會（NTUC）的CTC。CTC在公司層面將管理部門、工會和工作者聚集在一起，共同設計轉型計劃，把技術採用與工作重新設計、技能升級和更高的薪酬相配套。
讓我分享一個例子，總理在五一節集會上分享過。在淡港醫院（TTSH），醫療服務員工工會（HSEU）與管理部門通過CTC合作推出了智慧排班系統，可以處理多個班次模式，將排班時間從超過90分鐘減少到不足15分鐘，這樣護理經理可以將更多時間花在其他核心工作上。
資深註冊護士梁莉安（Lilian Teng），現年69歲，在淡港醫院工作了19年，她簡單地說：隨著技術讓工作對身體的要求降低，只要她保持健康，她就可以繼續有效地工作。這就是敬業勞動力的樣子。
但僅靠公司層面的努力是不夠的。隨著三方就業委員會的成立，政府、僱主和工會現在可以協調行業轉型，以工人為中心，確保人工智慧培訓、職位重新設計和轉型支援由基層塑造，而不僅僅是從上而下的決定。
至關重要的是，三方就業委員會可以升級CTC生態系統，將其範圍從大型僱主擴充套件到可能沒有資源獨自應對轉型的中小企業。CTC在公司層面與工人合作。三方就業委員會將在國家層面這樣做。他們共同確保人工智慧轉型不是強加給我們的工人，而是與他們一起進行的。用中文，議長。
（用普通話）：[請參閱原文演講。] 副議長先生，人工智慧已經來臨——而且來勢迅猛。但我們必須問：當人工智慧創造財富時，財富流入誰的口袋？我提出三點。
首先，公平分享蛋糕。人工智慧創造財富，但這些財富不能僅流向僱主。工人幫助烤蛋糕，他們也應該得到一份。
其次，如果失去工作，我們不能讓人們自生自滅。求職者支援的資格標準可以擴大，這樣更多的PMEs可以受益並獲得更大的保障。我們必須未雨綢繆——在暴風雨來臨前做好準備。
第三，一起走這條路。人工智慧轉型不能僅由僱主決定。俗話說，三個臭皮匠頂個諸葛亮。三方夥伴、工會和僱主必須共同努力，這樣我們才能走得遠、走得穩、走得快。
公平分享蛋糕，一起向前走。這是屬於每個新加坡人的人工智慧未來——一個以人民為核心的未來。
(英文):綜上所述，關於人工智慧是雙刃劍的言論已經很多。這個比喻雖然被過度使用，但這並非虛言。
人工智慧對廣大中層的影響意味著它既是一項'百年一遇的技術'，也可能是百年一遇的分化力量，將收益集中在控制人工智慧的人手中，同時卻導致那些幫助設計、實施和構建人工智慧的工人被取代。
公平增長、強化保護和充分參與的勞動力。這三項原則必須指引我們前進，如果我們希望在公平、韌性和為全體創造機會的框架下，在我們的人工智慧轉型中，將人工智慧確立為一個助力，因為每一位工人都很重要。副議長先生，我支援本動議。
副議長先生：政務部長Jasmin Lau
下午5時34分 數字發展與信息部政務部長（勞家蕙女士）：副議長先生，我今天認真傾聽了議員們的意見。議院內普遍對人工智慧對我們工作者的影響表示真誠關切。這些關切是真實的。政府與大家一樣有這樣的關切。
我們不能放慢人工智慧的發展。但我們不會讓其結果聽天由命。我們將努力促成在這裡繁榮發展的公司與為此付出努力的工作者之間達成不同的協議。
在新加坡運營和增長中獲益的公司，我們期待工作者獲得公平的協議。不僅僅停留在言辭上，而要體現在工作設計、人員培訓以及收益分享的方式上。在公共資源和政策用於支援業務轉型的地方，我們期待公司為工作者交付明確而有意義的成果。
在我與數字發展與信息部(MDDI)和教育部(MOE)的各種對話與合作，以及通過經濟戰略評審，相同的關切一次又一次地被提出。我的工作五年後還會存在嗎？人工智慧會加劇不平等，讓弱勢群體掉隊嗎？如果人工智慧提高了公司的生產力，工作者能分享這些收益嗎？
這些並非不合理的擔憂。這些擔憂來自於那些辛勤工作、多年積累技能和經驗、如今卻感受到腳下地面在變化的人們。我將逐一回答這些問題。
首先，關於今天的工作是否會繼續存在。讓我坦誠相待。有些職位將發生重大變化。那些主要建立在重複同樣步驟基礎上的職位最容易受到影響。這並不是對從事這類工作的人們價值的判斷。這對我們政府和僱主發出的訊號是，我們需要現在就採取行動，而不是等到破壞到來後才行動。我們將會採取行動。
但是，AI不僅僅是一種取代工作的技術進步。與此同時，它正在開創全新的工作方式和以前不存在的新型職位。
一些學者將AI描述為"發明發明方法的發明"。它擴大了可以解決的問題的範圍、可以構建的產品以及我們可以建立的行業。新加坡的一支小型生物科技團隊可以進行的實驗，在幾十年前需要國家級實驗室才能進行。一位單獨創始人可以釋出的軟體，在三年前需要一支百人規模的公司才能交付。
因此，競爭加劇，但前沿也向外擴充套件。這就是為什麼由高階議會秘書Goh Hanyan和我共同主持的經濟戰略審議委員會專注於確定新的領域，讓新加坡能夠利用人工智慧建立真正的競爭優勢。總理的國家人工智慧委員會將推進這項工作。
議員指出了人工智慧對中小企業的影響，因為人工智慧使日常和分析性任務自動化。經濟戰略審議團隊認識到這一點，這就是為什麼幫助企業和員工主動應對轉變成為了由國務部長Goh Pei Ming和國務部長Desmond Choo主持的委員會的重點。
對於失業工人，委員會研究了政府、僱主和工會如何能夠提供更及時的幫助。正如我們在中期更新中所提到的，經濟戰略審議正在研究鼓勵更早發出裁員通知的方法，這是由Mr Ng Chee Meng提出的。
關於中小企業(PME)具體地，委員會認識到他們可能面臨更大的就業不確定性，並將建議提供更有針對性的支援。這包括考慮加強《求職者支援計劃》(Jobseeker Support Scheme)，正如 Patrick Tay 先生所建議的那樣，以及汲取私營部門的專業知識來加強對該群體的安置支援。
對於面臨失業風險的工作者，《經濟戰略審查》將建議實際的方式幫助他們轉向更有彈性的、需求更強的職位。我們將確定勞動力需求持續且人工智慧失業風險較低的行業，並與這些行業的工會和僱主合作，為過渡中的工作者創造明確的、得到支援的進入點。我們必須使這些路徑可行，而不僅僅是可見的。
舉例來說，從事日常行政工作的中期職業工作者，例如從事資料輸入或客戶服務的工作者，可能會擔心人工智慧取代他。在工作便利化和再培訓支援的幫助下，該工作者應該能夠轉向具有建立在其現有技能基礎上的職位的行業。例如，該工作者可以探索醫療保健管理中的相鄰職位。由於人口增長和醫療保健需求的增加，我們在這裡看到強勁的需求，醫療保健需要獨特的人類技能，這些技能對中斷的抵抗力更強。
這一切需要的不僅僅是課程。它需要僱主、工會、培訓提供商和安置支援緊密合作，以便工作者在過渡期間不會掉隊。
世界上沒有哪個政府對這一轉變有所有的答案，我對任何聲稱相反的政府都會保持謹慎。新加坡能夠承諾的是：我們不會等待完美的解決方案後再採取行動。我們現在就開始，並將在進行過程中調整我們的努力。
其次，關於不平等。議員們的擔心是對的。能夠放大能力的技術也可能擴大那些快速適應者和那些努力跟上者之間的差距。正如 Mark Lee 先生所指出的，我們面臨的一些風險是生產率提高更多地流向那些已經領先的人，而職業階梯的底部可能面臨侵蝕。
我們的回應是提高底線，擴大入口。這意味著更早開始：在我們的學校中建立人工智慧素養，使所有學生都能對人工智慧充滿信心，而不僅僅是那些有機會獲得資源的學生。
目前，每個工業和技術教育學院(ITE)和理工學院學生已經被教授人工智慧素養作為他們課程的一部分，我們現在正在將人工智慧素養和安全的人工智慧工具帶入小學和中學課堂。這意味著所有學生，無論其經濟背景如何，都可以安全地學習關於人工智慧的知識。他們還可以學習人工智慧如何能夠有利於他們的學習，例如幫助他們完善他們的想法，他們也學習何時不應該使用人工智慧。
正如 Desmond 部長今天早些時候所指出的那樣，我們致力於支援可能缺乏強有力的家庭或父母監督和支援的學生。雖然學校中的人工智慧素養將為他們提供良好而堅實的基礎，但我們必須繼續與社群和自助團體發展夥伴關係，以確保學校外部的監督和支援繼續。
學習必須在畢業後繼續。從2026年下半年開始，我們所有的高等教育機構(IHL)將為其校友提供精選的人工智慧相關課程，有重大折扣，期限為一年。
對於已在勞動力隊伍中的工作者，完成精選人工智慧培訓課程的新加坡人將獲得六個月的高階人工智慧工具的免費訪問。我們將跟蹤採用和使用情況，以檢視我們是否需要做更多。
每個新加坡人，無論其起點如何，都應該有機會嘗試人工智慧工具並建立熟練度。
第三個問題是最難的，也是最重要的。工作者是否會分享收益？我們應該說清楚：這不會自動發生。僅憑技術本身會導致非常不平衡的結果。這就是為什麼這不僅僅是一個市場問題。這是我們如何塑造經濟中的規範和期望的問題。所以，讓我清楚地說明我們的期望。
從人工智慧中受益的公司應該投資於他們的人員，而不僅僅是技術。這意味著儘可能地培訓現有工作者，而不僅僅是僱用新工作者。這意味著促進員工使用前沿人工智慧工具的途徑，建立實踐社群，激勵學習和提升技能。
它還意味著在與工作者密切磋商的情況下重新設計工作，正如 Yeo Wan Ling 女士所建議的那樣，使人們能夠與人工智慧一起工作，使用判斷力、背景和經驗，而不是簡單地將工作者視為要減少的成本。當職位確實改變或消失時，它意味著在訴諸裁員之前，認真努力在組織內重新部署和重新培訓工作者。
我們不僅僅是在要求我們的公司進行國家服務。我們要求他們做的是他們自己長期利益所在的事情。在人工智慧時代，人類的直覺和直觀性仍然是關鍵。我們都知道，當我們使用人工智慧時，我們需要引導它。在我們反覆完善輸出時，提出正確的問題並運用判斷力。
這不是一次性的。如果你不開發理解你的組織背景的人，並使用這種知識來加強你的人工智慧系統，你將來會被遺留一個非常淺薄和空心的公司。如果這裡的公司用人工智慧完全取代人類，他們將來會發現自己沒有競爭優勢，因為人工智慧對所有公司都是可用的。他們還會發現自己受到人工智慧公司的擺佈。所以，我們的目標是找到一種方法，使我們的公司在長期內最好地為可持續增長而定位。
Saktiandi Supaat 先生提出了需要平衡的監管方法，這些方法不會阻礙人工智慧的採用。確實，我們不會試圖通過立法來獲得良好的結果。這從來都不是新加坡的主要方法。但我們同樣清楚，"自願"不能意味著"在實踐中是可選的"。
在部署公共資源的地方，我們將要求工作者成果。我們將與公司合作以滿足這些期望。如果存在持久的差距，我們將審查我們的支援如何適用。我們將與三方夥伴討論如何公平有效地做到這一點，以鼓勵公司投資於培訓、工作重新設計、重新部署和安置。
如果我們做得好，我們將能夠在人工智慧時代創造和維持良好的工作。好工作不僅僅是存在的工作。它是允許工作者進步的工作。它應該公平支付，反映技術帶來的生產率提高。它應該建立仍然相關的技能，包括作為常規在職培訓的一部分，以便工作者不會被困在容易被自動化替換的任務中。它應該給工作者一種尊嚴和代理權感，而不是將他們的角色簡化為遵循機器生成的指示。
我們已經看到，當有強有力的承諾時，這是可能的。在 PSA，人工智慧和自動化幫助實現了創紀錄的貨物量。與此同時，該公司對2,000多名工作者進行了再培訓並重新部署到高技能職位。他們繼續僱用數千名更多員工，因為他們的增長速度快於競爭對手。
對 Andre Low 先生，我要說：自動化和增強並不相互排斥。保護一名工作者可能意味著有意自動化重複和體力上要求很高的任務，並升級同一工作者的技能，以便技術可以在他承擔更高價值的角色時增強他的能力。
即使是較小的企業也在發揮他們的作用。以我們本地的當鋪 Maxi-Cash 為例。過去，希望交換珠寶的客戶會與銷售顧問互動，銷售顧問會將他們的案例轉交給評估員以評估珠寶的真實性。Maxi-Cash 通過對25名銷售顧問進行再培訓以使用人工智慧啟用的認證系統來增強此過程，該系統可以在僅五秒內準確評估珠寶的成分。現在，這些銷售顧問可以補充現有的評估員庫，減輕他們的工作量並減少客戶等待時間。這是我們希望在新加坡看到的作為常規而非例外的負責任轉變。
副議長先生，我認真聆聽了今天在議院中從兩方議員分享的許多建議和觀點。我們可能在具體政策思想或特定措施應如何設計、融資或排序上有所不同。這是民主辯論的性質，這是非常健康的。
但我相信這個議院中對一項基本原則有廣泛共識：增長和進步的收益必須與所有新加坡人公平而廣泛地分享。這不應該是政黨或意識形態問題。這是我們作為新加坡人必須共同堅持的原則。
所以，讓我明確地說這一點。如果新加坡在我們的人工智慧雄心中成功 - 而我們應該永遠不要假設成功是自動的，因為它將需要持續的努力、艱難的選擇、調整，也許還需要一些好運。但如果我們成功，那麼政府將確保收益被廣泛分享。
收益不能僅僅流向那些已經擁有資本、優勢或訪問許可權的人。它們必須轉化為更好的工資、更好的機會和所有新加坡人的更大安全性。對工人的最好保護不僅僅是中斷後的再分配。它是從一開始就塑造如何建立和分享收益，並確保新加坡工作者在人工智慧經濟中保留代理權。
這個政府在新加坡數十年的發展中能夠取得這些成果。我們決心在應對這一人工智慧轉變時繼續這樣做。我們的政策從未是靜態的。隨著情況的變化，我們已經持續調整、重新整理和加強了它們。這種紀律將繼續。
最終，每個新加坡人都應該能夠看看新加坡所建立的東西並說，"我在這一進步中有利益關係。我在這一增長中有份額。這個未來，也屬於我和我的家人。"
這種共同的承諾也是使新加坡對這一轉變的方法具有獨特之處的原因。我們的力量不僅僅是技術。這是我們跨越政府、企業和工會合作的方式。
對於正在觀看這場辯論的工人們，我想直接對你們說：政府站在你們這一邊，我們正在搶在中斷影響你們之前採取行動，而不是之後。你們不會獨自面對這一切。我們今天在這個議院做出的承諾，就是對你們的承諾。
對我們的企業領導者：人工智慧為你們提供了強大的新能力。但你們如何使用這些能力將決定你們公司的未來，以及你們與那些與你們共同建立公司的人之間的關係。在10年內領先的公司不是那些削減成本最快的公司，而是那些通過將人的判斷與機器能力相結合來建立更強大團隊的公司。
但我也想澄清另一件事。不是每個企業都需要採用人工智慧，也不是每個追求都需要從人工智慧轉型的角度來看待。完全由人創造的東西有真實的價值，隨著人工智慧變得更加普遍，這種價值可能會增長，而不是減少。
當我們周圍的一切都是自動生成、最佳化和擴充套件的，那些不是這樣的東西就會脫穎而出。無法重複的現場表演和返場。承載人類手工痕跡的手工陶瓷碗。用心和工藝精心準備的食物，而不僅僅是一致性。與花了一生時間打磨藝術的書法大師的對話。
我認為我們將看到對這些東西的欣賞重新復興。新加坡不僅應該認可這一點，我們應該擁抱它。我們的工匠、表演者、工藝師不是在逆潮流而行。在充滿人工智慧生成內容的世界中，他們可能會發現自己恰好在世界所注視的地方。
超越近期的過渡，有一個長期的問題我們必須回答。我們現在需要對教育系統做什麼，為我們的學生準備未來世界？
我們必須接受人工智慧將在機器擅長的任務中繼續進步。因此，我們更需要專注於使我們獨特人性的東西。提出沒有人想到過的問題的好奇心。以訓練資料無法預測的方式連線不同領域思想的創意。能夠讀懂氛圍、贏得信任、知道什麼時候最有效的解決方案可能不是正確解決方案的同理心。
我們經常稱這些為軟技能。在人工智慧時代，它們將成為我們人民和新加坡競爭優勢的硬核邊緣。這就是為什麼我們將審視我們的教育系統，以確保我們以與一直以來對學術卓越所應用的相同嚴謹性和意圖來發展這些品質。
我們必須繼續建立堅實的基礎，確保我們的學生不會過度依賴人工智慧的快捷方式。我們的人腦是需要鍛鍊的肌肉，真正的掌握——那種能在壓力下堅持、人工智慧無法簡單替代的掌握——來自辛勤工作、實踐和深入理解。因此，聽到陳慧玲女士同意這一點很好，我們感謝她支援我們的方法。
但嚴謹和探索不是對立的。真正掌握某件事的學生正是有信心超越它的人。他會提出更難的問題，承擔沒有明顯答案的問題，他會發展出真正屬於自己的興趣。我們正在努力建立的是一個既要求深入的學科紀律，又要求廣泛自由的教育系統。不僅僅是因為我們的學生值得擁有兩者，而是因為新加坡的未來取決於兩者。
這並不意味著放棄我們的標準。這意味著擴充套件我們認為卓越的範圍。提出意想不到問題的學生，出於真誠興趣深入追求某事的學生，能夠持有兩個矛盾觀點並逐一處理的學生——這樣的學生並未落後。在我們正在建立的世界中，這樣的學生可能走在我們所有人的前面。
我們致力於與我們的教育工作者、家長和年輕的新加坡人一起做這件事。因為如果我們把這件事做對，如果我們培養一代不僅懂人工智慧而且深深具有人性的人，那麼新加坡不僅會度過這一過渡期。我們將成為下一個人類進步時代圍繞其構建的社會。
副議長先生，我們今天對這一過渡將要求什麼——對政府、企業和工人——都是誠實的。並非每條路都會順利。有些人將面臨真正的中斷，我們的責任是確保沒有人獨自面對它。我們將使人工智慧為新加坡人服務。我們將確保隨著經濟增長，工人與之同步前進。
但我想以我相信我們的注意力最終必須停留的地方結束——我們正在建立的一代。如果我們培養出好奇、富有創意、深深具有人性的新加坡人，他們能提出機器無法提出的問題，贏得演算法永遠無法贏得的信任，那麼我們不僅僅是管理這一過渡。我們將定義它之後的樣子。我支援這項議案。[掌聲。]
副議長先生：國務政務部長陳德勇。
下午6時 總理公署國務政務部長（陳德勇先生）：副議長先生，我首先宣告我作為全國職工總會副秘書長的權益和新加坡工業及服務業僱員聯合會(SISEU)執行秘書的權益，在這些職位上我密切參與支援我們的工人。今天，我想繼續支援我們的資深工人，並反映他們在人工智慧時代的關切。
讓我先分享一個關於傅女士的故事，她是一位55歲的求職者，向全國職工總會的e2i尋求幫助。在離開她的上一份工作——超過20年的工作——後，她發現求職過程發生了相當大的變化。甚至簡歷寫作也改變了。簡歷過去是為人寫的，但她發現如今它們經常首先被機器篩選。工作申請也轉移到了不那麼直觀的數字門戶，不太容易讓像她這樣的人來導航。所以，她感到迷茫和不確定。
傅女士的經歷在資深工人中並不罕見，反映了他們對工作流程變化的焦慮。對於一些人來說，人工智慧提供了真正的機會。對於其他人來說，它造成了不確定性和焦慮。對於我們許多資深工人來說，他們的經歷由三個關鍵差距塑造，我認為我們集體必須努力解決。
首先是接入差距。雖然新加坡在縮小資深人士的接入差距方面取得了進展，但他們中的智慧手機所有權仍然落後於其他年齡組。此外，資深人士可能對人工智慧工具的接入較少。
我在我的工會SISEU組織的一次人工智慧研討會中親眼目睹了這一點，大約有90名工會領導人參加，他們的平均年齡約為53歲。雖然他們都使用智慧手機，但許多人是第一次嘗試人工智慧工具。可以理解的是，最初有些猶豫。但通過簡單、有指導的用例，他們很快上手了，並使用人工智慧為工會的家庭日和會員制活動生成海報。有些人甚至為他們的家人做生日和週年紀念橫幅。
當我們結束會議時，許多領導人主動與我分享他們享受了這次會議，現在他們意識到學習使用人工智慧工具並不那麼困難。他們唯一要求的是希望會議能更長，幻燈片上的字型對他們來說能更大。
這鼓勵了我，因為它表明問題不在於缺乏意願，而是缺乏接入機會，在某種程度上對於他們中的一些人來說，這是關於信心的問題，我們可以通過精心策劃和定製人工智慧工具的接入，為資深人士騰出時間學習和增加知識來克服這個問題。
其次，技能差距。我們看到培訓參與中的差距，人力部2025年報告發現，50至64歲的居民的培訓參與率最低，為44.5%，相比之下，40歲以下的人參與率約為60%。
告訴資深人士去提升技能、去參加課程和重新技能化很容易，但實際上，我們知道，面對承諾、賬單要付、時間和精力有限的情況，邁出第一步有時並不容易。我們必須記住，他們中的許多人已經經歷了多個變革和轉型週期，可能感到疲勞、不確定，甚至質疑更多培訓的相關性。
這些關切是真實的。我們必須以適當的速度通過實用和精簡模組為他們提供更容易獲得的培訓，並使人工智慧與他們的工作技能更加相關。
第三是機會差距。即使資深工人願意學習，他們在實際工作中可能沒有相同的機會從人工智慧中受益。
經合組織2025年就業展望強調，在經合組織國家中，通過實踐學習的機會隨年齡下降，62%的25至29歲的成年人報告有此類機會，但在60歲及以上的人群中降至45%。此外，皮尤研究中心2025年報告發現，在工作中使用人工智慧的工人中，73%的年齡在18至49歲之間，只有27%的年齡在50歲及以上。
這就是為什麼我們需要與僱主合作，為我們的資深人士提供使用人工智慧工具的機會，並在他們的工作中獲得生產率收益。資深工人帶來寶貴的經驗，藉助人工智慧，這些優勢可以更進一步。我在這個議院已經不少於兩次談過這個話題。
斯坦福數字經濟實驗室的研究支援這一觀點。它強調，在人工智慧的引入下，美國的資深就業一直保持韌性，甚至可能有所增長，因為他們看到資深人士帶來的隱性經驗、知識和軟技能使他們能夠通過人工智慧提高生產率。
對我們的資深工人，我們理解你們的挑戰，我們在這一過渡中與你們同在。與我們邁出第一步。開始參加課程，嘗試工具，從你們周圍的人那裡學習，你們就能在人工智慧經濟中茁壯成長。
副議長先生，本院現審議的動議很重要。該動議強調經濟增長必須以公平、韌性和全民機遇為基礎，並決心幫助工人把握這些新機遇。這一點也必須同樣適用於我們的資深工人。
NTUC認可，長期影響力最好通過夥伴合作模式來實現，並通過人工智慧就緒新加坡倡議採取了積極步驟，該倡議重點關注三個關鍵領域。
首先，培訓和提升工人技能，以填補技能差距和可及性差距。為了縮小技能差距，NTUC LearningHub開發了綜合的人工智慧學習路徑，針對不同熟練程度的學習者提供三個不同級別：基礎培訓，用於建立人工智慧素養和應用能力；中級培訓，針對特定部門或職位量身定製；以及針對從事深度技術且希望深化人工智慧專業化和能力的人員的高階培訓。
我很高興地注意到，迄今為止反響熱烈。自2026年2月以來，超過4,000名工人已報名參加LearningHub的人工智慧課程。我也非常高興，其中39%的學員是資深工人。
Neo博士建議對能力進行認證。這是LearningHub已經在做的事情。例如，它與公司密切合作，設計與政府技能框架一致且符合公司需求的人工智慧課程。我們也與AWS和Microsoft等行業領袖合作，根據行業需求對學員的能力進行認證。我們將繼續把這一做法擴充套件到更多部門和行業。
同時，在人工智慧就緒新加坡倡議下，我們也通過NTUC的工會培訓援助計劃（UTAP）為工會成員提供最高達50%的人工智慧高階訂閱工具補貼，以縮小可及性差距。我也很高興向各位議員通報，在這首批人工智慧工具訂閱中，NTUC的高階補貼確實涵蓋一系列工具，包括編碼和基於代理的工具，如Claude Code、Codex、Manus等。我認為總共有20或21個工具。
我們之所以將其作為會員特權提供，僅僅是因為我們使用現有的專為我們會員培訓設計的UTAP資金模式。但我們將繼續審查這一點，取決於未來的採納情況和興趣程度。
我們也與新加坡民航局等部門機構合作，為我們的工會領導人開發部門級人工智慧培訓路徑。
其次，我們通過NTUC的CTC支援企業進行業務轉型和職位重新設計。迄今為止，NTUC已建立3,800個CTC，啟動了900多個業務轉型專案，惠及超過300,000名工人。
讓我分享另一個例子。我知道在這次辯論中各位聽過許多例子。這個例子來自常綠集團，一家本地辦公用品和文具供應商。通過與新加坡手工及商業工人工會（SMMWU）的CTC補助金專案，它實施了一個由人工智慧驅動的電子訂購系統，以自動化訂購流程並改進庫存管理。有了這個新系統，訂單處理等手工工作減少了約60%。工人可以專注於更高價值的任務，如管理客戶關係和利用資料最佳化庫存。
隨著生產力的提高，該公司能夠處理40%更多的訂單，併為員工提供加薪。這就是我們所說的互利共贏的成果——企業生產力提高，我們的工人也隨之進步。
第三，我們通過新產品和服務改進職位匹配，幫助工人獲得好工作。讓我回到我演講開始時提到的Foo女士。在e2i職業教練的密切支援下，她對自己的技能和新的職業市場有了更深入的瞭解。她的教練還向她介紹了NTUC的人工智慧職業教練和e2i的人工智慧面試官。在支援和鼓勵下，Foo女士能夠自信地使用這些人工智慧工具來完善簡歷，並在實際面試前進行面試練習。我很高興分享，Foo女士已找到新的職位，並在此過程中增進了對人工智慧的瞭解。
副議長先生，人工智慧就緒新加坡倡議是我們如何實現本動議意圖的一個例子。它支援工人建立人工智慧技能，為他們提供運用這些技能的工具，並與企業合作以提高生產力和創造新機遇。
展望未來，我們歡迎公司和合作夥伴參與，隨著我們擴大努力和擴大覆蓋範圍，最終為我們的企業和工人提供更好的支援。
副議長先生，這些努力很重要，我們已經看到令人鼓舞的成果。但人工智慧帶來的變化規模也很大，工人中存在很大的不確定性和焦慮。這就是為什麼三角合作制——其建立在數十年開放溝通和信任基礎上的良好記錄——對應對這一挑戰至關重要。
新加坡以前曾成功應對過重大變革。在1980年代，當計算機首次進入工作場所時，工人們擔心這些機器可能會取代他們在資料錄入和檔案管理中的角色，而公司則擔心成本、技能短缺以及日常運營中斷。
但三方夥伴積極行動。政府投資基礎設施和技能培養，包括國家電腦局，以推動全國採用資訊科技。僱主主動進行業務轉變，採用新技術和重新設計的工作流程。勞工運動推動大規模技能升級，組織研討會和講座，為工人們在心理和實踐上做好應對變革的準備。
由於三方夥伴團結一致地行動，企業的生產力提高了，工人從事了薪資更高的更好工作，新加坡增強了競爭力。
副議長先生，三方就業理事會將是實現我們對AI時代的共同願望的關鍵平臺，正如本動議中所列明的。它將建立在政府、僱主和勞工運動各方的努力基礎之上，使夥伴能夠擴大外展範圍、加快政策實施、調配資源，以便工人和企業能夠從人工智慧中抓住機遇。
三方就業理事會將採取務實和迭代的方法。我們一開始可能沒有所有答案，但我們確信，在數十年的合作和共同目標基礎上建立的深厚信任將使我們能夠實現包容性經濟增長的願望。副議長先生，我現在將用普通話發言。
（以普通話發言）：【請參閱講話原文。】副議長先生，新加坡需要利用人工智慧來推動下一階段的經濟增長。但更重要的是，這種增長必須建立在對所有人都公平、包容和有韌性的基礎之上。
新加坡的三方夥伴——工會、僱主和政府——多年來緊密合作，建立了深厚的信任，使我們能夠團結一致、共同克服挑戰，無論面臨什麼困難。我們將繼續支援我們的工人在人工智慧時代抓住機遇並增強他們的競爭力。
今年2月，全國職工總會推出了'AI就緒新加坡'倡議，積極鼓勵工人學習和掌握人工智慧工具。該倡議幫助他們彌補意識、技能和獲取方面的差距，使他們為人工智慧驅動的經濟做好更充分的準備。正如諺語所說，'機遇偏愛有準備的人，成功屬於最堅持的人。'人工智慧時代已經來臨。我希望每個人都與全國職工總會一起，積極提升技能、學習並應用人工智慧。
(英文):副議長先生,人工智慧是我們這一代人的決定性技術。但我們面臨這一挑戰時,擁有通過數十年三方合作建立的堅實基礎。
對於我們的三方夥伴,讓我們繼續緊密合作,幫助我們的企業轉型並保持競爭力,同時支援我們的員工保持生產力並充分利用機會。
對於我們的員工,我們將繼續支援你們利用人工智慧。
對於我們的資深員工,你們的經驗很重要,我堅定相信這將是人工智慧時代的優勢。與我們一起邁出這一步,提升技能並不斷學習,包括向你們年輕的同事學習,我們可以共同努力彌合接入差距、縮小技能差距,為所有人創造更多機會。因為在新加坡,我們一直相信進步必須具有包容性,當我們向前邁進時,我們一起邁進,因為每一位員工都很重要。副議長先生,我支援這項動議。[掌聲。]
副議長:人力部長陳思倫。
下午6時16分 人力部長(陳思倫醫生):副議長先生,尊敬的先生,讓我首先承認許多新加坡人現在正在感受的——不確定性、焦慮、一種腳下地面在移動的感覺;世界變得不如過去那麼可預測,貿易緊張局勢、供應鏈脆弱性、中東地區的戰爭和油價的急劇上升。
在我們本地,本議院議員已經談到困擾所有同胞心頭的事情:人工智慧可能削弱我們的技能、我們多年積累的經驗,甚至可能奪去我們的工作的焦慮。這種焦慮因大型科技公司宣佈因人工智慧採用而裁員的新聞而進一步加劇。
這些是合理的關切，我們認真對待。這樣規模的變化確實令人不安。但人工智慧能夠並將創造我們目前無法完全想象的機會。當然，與此同時，它也會帶來我們無法完全預見的擾亂。
但與此同時，出現了一些早期跡象給我們理由保持謹慎樂觀。最近的全球調查顯示，三分之二進行過早期人工智慧驅動裁員的公司已經開始重新招聘。為什麼會這樣？因為他們發現人工智慧可以處理可預測的和常規性的工作，但客戶仍然需要人類的判斷力、同理心和真實的聯絡，這些是人工智慧無法提供的。
讓我舉一個個人的小例子。在準備今天的演講時，我的團隊使用人工智慧來幫助完善我的工作。它提供了有用的參考資料，包括我們人力部最近釋出的研究，顯示新加坡只有約6%的公司因採用人工智慧而減少了員工數量。
但有一件事它無法理解——許多許多工人所感受到的影響和焦慮。它無法提供同情，無法同情人，無法理解細微差別，也無法生成能夠捕捉工人真實經歷本質的政策回應。這是任何演算法都無法替代的。
[議長主持]
擺在我們面前的動議做出了四項承諾。政府和人力部認真對待每一項——都是作為繼續建設和進一步發展的基礎。
黃志明先生談到了人工智慧對我們勞動力的變革性影響。政府長期以來一直認識到人工智慧的潛力。我們目前的努力建立在這一領域已經完成的工作的堅實基礎上。我們在2019年制定了第一個National AI Strategy，遠早於ChatGPT的推出，我們在教育、醫療、物流、安全和市政服務等領域啟動了國家人工智慧專案。
當大型語言模型在2022年末爆炸性出現，使人工智慧變得易於獲取和通用時，我們在2023年以National AI Strategy 2.0重新整理了我們的戰略，並制定了投資超過10億美元用於人工智慧計算能力、人才和產業發展的計劃。這包括建立AI卓越中心和增加人工智慧從業者的數量。
隨著人工智慧加快步伐並與我們外部環境的重大轉變相互作用，我們去年召集了Economic Strategy Review來加強我們的應對。最近在今年的預算演講中，我們成立了由總理主持的National AI Council，以推動使用人工智慧對我們經濟的實際轉變。
在每一步，我們都與我們的三方夥伴積極合作，以在各個部門推動具體行動和轉變。因此，雖然我們將走入不可預知的水域和不確定的未來，但我們可以充滿信心地這樣做，我們並非完全毫無準備。
各議員都對人工智慧對就業流失的影響提出了關切，許多人也提出了有思想性的建議，說明我們如何可以更好地在這一過渡期間支援工人和企業。我們聽到了你們的關切，我們歡迎來自議院兩邊議員提出的建議。
事實上，議院中對我們試圖實現的目標存在廣泛的共識，即在這一人工智慧轉變中實現全民包容性增長。我們可能不同的地方在於我們如何實現這一目標。我們的方式一直是投資於我們的人民，保持我們工人的經濟價值，並塑造人工智慧收益如何被創造和分享。我們不想停留在恐懼和憂慮上，我們想要能夠激勵和鼓舞我們的勞動力繼續增長。
Mr Gerald Giam先生和Mr Andre Low先生強調了對包容性增長的結構性威脅。我讚賞他們在這個問題上認真的態度。我們認識到這些困境。問題是：在哪裡以及如何進行干預？
Giam先生提議建立National AI Equity Fund，用從受益於人工智慧的公司的資金向每個新加坡公民支付500美元。羅先生同樣提議通過冗餘保險對那些失業的人進行賠償。我認識到需要加強我們的系統，以確保在這一過渡期間沒有人會掉隊。我同意生產率收益的廣泛分享不會自動發生，因為市場單獨無法保證良好的社會成果。
讓我明確說明。政府一直知道這一點，一直在為此採取行動。Giam先生和羅先生的提議都基於一個更悲觀的前提，即新加坡人在人工智慧轉變中本質上是被動的乘客，沒有能力抓住機會，只能依靠支援來應對他們無法掌控的旅程。
我不能接受，也不會接受這樣的前提。你們的提議都不是賦權。對我來說，這是妥協。認命於大規模失業是不可避免的這一事實，我們能做的最好的事就是緩解打擊。我們應該更加相信我們新加坡同胞的韌性和適應能力。
如果工人被排除在經濟之外，僅靠再分配是不夠的。新加坡的傳統一直是投資於人而不是為他們補償困難；這是我們真正的政策傳統，而不是羅議員所描述的那樣。
更好地使用人工智慧採用產生的任何盈餘的方式是資助可獲得和有效的技能提升，以增強新加坡人的價值。為此，政府在過去五年中在本地勞動力倡議上花費了超過$10 billion。
擺在我們面前的選擇，議院議員們，是在兩個非常不同的願景之間進行選擇。一個說你得到救濟，然後，用那個，得到機器生產的餡餅的一小份。另一方面，我們認為你應該與機器一起做大餡餅，通過好工作和好工資分享我們的經濟繁榮。
第一個願景初看起來可能很慷慨，但最終，會限制和削弱你更廣泛的目標。第二個願景對政府、僱主和工人都提出了更高的要求，但它將我們所有的新加坡同胞視為有能力的成年人，他們的未來值得投資，而不是一個需要通過轉移支付來管理的人口。
議院議員們，我相信第二個願景是可能的，因為當人工智慧改變我們的工作方式時，有些工作會演變。有些工作可能會消失，但如果我們能讓每個人都在同一艘船上一起移動，我相信我們會成功。正如Mark Lee先生和Yeo Wan Ling女士所說，它也為企業和工人創造了新的機會。我們的責任和重點是幫助我們所有的工人和企業抓住這些機會。
因此，我們在這場辯論中永遠不應該基於焦慮和憂慮來建立論點。我們的方法不是害怕未來，而是讓我們自己塑造未來，因為這是真正的新加坡精神。
我們已經用這種精神度過了每一波技術和經濟重組，但我們並不自滿。人力部正在密切監測人工智慧對我們勞動力的影響。我們首次對企業的調查顯示，人工智慧目前在新加坡增強而不是替代勞動力。目前約十分之三的公司已經採用人工智慧，而在已經採用人工智慧的公司中，只有少數，約6%，報告了員工數量減少。
更常見的是，企業正在重新設計工作，他們正在建立新的人工智慧相關角色，表明人工智慧正在改變工作方式、工作如何被重新組織，而不是減少工作。十分之七的使用人工智慧的企業已經看到了生產率的提高。
然而，正如我所說，我們永遠不應該自滿。我們必須為人工智慧採用步伐加快、勢頭增強、規模擴大時，對工作的影響會更大做好準備。這就是為什麼我們不斷地為自己做準備。
我們的目標是幫助更多企業成功。同時，讓他們的員工利用人工智慧做得更好，而不是被人工智慧取代。讓他們的工作變得更有成就感、更有意義，而不是相反。以及人工智慧的好處在企業本身和勞動力之間分享。
對於可以以更靈活節奏工作的工人，人工智慧可以實現新形式的靈活工作和由小團隊甚至"solopreneurs"完成的兼職工作。除了靈活性外，人工智慧還可以重塑誰參與我們的勞動力，包括老年人，正如Poh Li San女士所談到的。我們將通過三方老年就業工作組探索如何擴充套件靈活工作模式。
Hamid Razak博士談到了他從基層聽到的猶豫和焦慮，特別是來自年長的PMEs，他們想知道他們的技能是否仍然有一席之地。Yip Hon Weng先生也要求為企業提供更好的支援。讓我分享政府在做什麼來為個人和企業為這一轉變做準備。
首先，我們正在改革我們的勞動力和技能支援系統，以提供更及時和有效的支援。正如我在昨天關於《技能和勞動力發展局法案》二讀五小時辯論後分享的那樣，這個SWDA的成立將把SkillsFuture Singapore和Workforce Singapore的技能和就業便利能力集中在一個屋簷下，使個人和僱主獲得適當支援變得更加無縫、更加綜合。
我們同意黃志明先生的觀點，即從SWDA中獲取的所有資料所產生的情報必須繼續建立在信任的基礎之上，我們期待與三方夥伴密切合作，確保我們對勞動力市場的評估繼續是有根據且最新的。這將是我們如何走在破壞之前並支援由人工智慧變化驅動的工人的重要組成部分。
這包括面臨自主駕駛車輛部署的平臺工人，正如楊玉玲女士所強調的那樣。人力部和新加坡勞動力發展局已經與運輸部和三方夥伴緊密合作，為這些司機加強轉型過渡路徑，以應對即將到來的自主駕駛車輛部署。我想補充的是，實際上作為代理機構，是新加坡勞動力發展局。但實際上，目前與運輸部緊密合作的是SkillsFuture新加坡公司和新加坡勞動力發展局。
第二，我們將採取更多措施提高新加坡人的人工智慧素養。目前，MySkillsFuture網站上有超過1600門與人工智慧相關的課程。我們將推出診斷工具，供個人評估其當前的人工智慧準備程度，並找到適合其需求的課程，提供與僱主需求相一致且經驗證的培訓成果。
從今年下半年開始，註冊參加選定SkillsFuture人工智慧課程的新加坡人將獲得六個月的高階人工智慧工具免費使用權。這將幫助他們將課堂學習應用到日常生活和工作中。鄧慶康先生建議不設條件地使所有人都能獲得此訪問許可權。這是政府仔細考慮過的事項。但並非所有新加坡人都需要前沿的智慧體級工具。對許多人來說，免費版本已足夠，而且應用廣泛。
通過將補貼與培訓掛鉤，我們能夠更好地針對那些更認真地尋求提升人工智慧使用水平的人，並幫助他們以最優和負責任的方式使用這些強大工具。正如何錦賢副教授和楊瑜先生之前所分享的那樣，我們希望新加坡人利用可用資源，在學習之旅中積極主動。
正如劉美娟政務部長所分享的，新加坡資訊通訊媒體發展局也將擴充套件TechSkills加速器專案，開發人工智慧雙語工作者，首先從會計、法律和人力資源專業人士開始。[請參閱《人力部長的澄清》，官方報告，2026年5月6日，第96卷，第30期，書面宣告更正部分。]更多細節將在適當時候公佈。
第三，為了支援企業，我一再強調我們已為企業勞動力轉型計劃撥出超過4億美元。我不想過多贅述，因為我相信我在昨天的二讀演講中已經涵蓋了大部分內容。但葉先生問這些補助金是否可以與勞動力成果條件掛鉤。今天，利用勞動力發展補助金(職位重新設計+)的企業需要在其轉型計劃中支援勞動力成果，如工資增長和員工留任。今年晚些時候，符合條件的企業還將在重新設計的SkillsFuture企業信用下獲得1萬美元，這可用於抵消符合條件的勞動力轉型專案的自付費用，包括企業勞動力轉型計劃下的專案。
我們同意李明先生的觀點，認為行業協會和企業勞動力轉型計劃可以解決這些問題，同時確保勞動力在此過程中被納入其中。商會在將企業與正確的專業知識和資源相連線方面發揮特別重要的作用。這就是為什麼我們任命新加坡商業聯合會和SNEF作為企業勞動力轉型計劃的錨點專案夥伴，以便將綜合勞動力轉型支援直接帶給企業，並幫助我們加快各行業的人工智慧採納。
我們也在支援勞動運動在企業和工人轉型中的努力。政府在2025年向新加坡全國工會聯合會合作轉型中心贈款增撥約2億美元，並將贈款延長至2028年。最近，我們與新加坡全國工會聯合會合作擴充套件贈款，以更好地支援蜂王企業推動叢集級轉型。
儘管我曾短暫離開去接聽電話，但聽到黃志明先生分享合作轉型中心如何幫助許多企業實現轉型，同時改善工人生活的情況時，我感到欣慰。我特別注意到他的敦促以及他進一步擴充套件合作轉型中心倡議的建議，以及他將合作轉型中心提升到三方水平的雄心。我們期待與三方夥伴合作，共同探索實現這一目標的途徑。
有人呼籲我們超越專案層面的干預措施，做出更具結構性的轉變，改變企業投資工人的財政激勵。結構性機制，如洛和民先生所呼籲的那樣，已經存在。像SkillsFuture企業信用這樣的補助金為企業投資其員工能力創造了直接的財政激勵。我們將繼續審查和加強此類支援，作為新加坡勞動力發展局工作的一部分。
我認為我們所有人都應該認識到支援企業轉型的重要性，不是籠統的、大規模的戰略，而是差異化的、精準的、按行業和按企業分別進行的企業轉型支援。儘管這更加繁瑣，但我相信從長遠來看，這也更具可持續性。
最後，我們加強了對失業工人的轉型支援，使他們能夠更強勁地反彈。政府無法保護每一份工作，但我們肯定會盡最大努力支援和保護每一位工人，因為每一位工人都很重要。
因此，隨著人工智慧轉型，工作流程將重新組織和改變，工作也將改變，某些工作可能會被替代。經歷轉型可能很具挑戰性。但我向所有工人保證，你們不會孤單。
我們已經認識到，隨著變化步伐加快，我們必須加強支援機制。這就是為什麼我們去年推出了SkillsFuture求職者支援計劃。這是我們在《前進新加坡》下重新整理的社會契約的一部分。該計劃向非自願失業人員提供臨時財政救濟和求職支援。它對許多新加坡人產生了影響，幫助他們重新站穩腳跟，以信心重返工作崗位。
洛和民先生可能對該計劃有些誤解。它不是冗餘保險，因為它不僅僅是為失業提供現金賠付。它是對再就業的支援。該計劃在工人的再就業之旅中提供支援。它為低收入和中等收入群體提供一定程度的財政支援，正是為了他們不會急忙接受第一份可能不適合的可用工作。
新加坡勞動力發展局用實際的全面支援補充求職者支援計劃，以提高其求職的質量。我們意識到長期失業會損害工人的長期職業前景，這就是為什麼財政支援有時間限制。它逐步下降，因為我們相信工人非自願失業的前兩到三個月是影響最大的時期。因此，我們在初期提高支援水平以鼓勵工人，提供支援提升，當逐步下降時，我們希望工人能夠找到適合他們的工作。
但我們聽到一些呼聲。黃志明先生和泰先生提議提高JS計劃收入門檻，以更好地支援高收入人群。我們將研究該計劃如何改進，並將仔細研究這一點。
我們也聽到黃志明先生關於要求政府提前通知遣散的呼籲，即在員工最後工作日之前，以及李明先生對企業在這方面的顧慮的反思。我們希望取得正確的平衡。三方夥伴已經在正在進行的《就業法》審查中討論縮短遣散通知期限。我們希望看到向政府的通知儘可能在受影響工人最後工作日的前後發生，因為這樣還能使及時的就業促進支援工人成為可能。
關於鄧慶康先生加強對失業工人保護的建議，《就業法》已經通過建立程式保障（如通知期和爭議解決途徑）提供了基於廣泛基礎的保護。這適用於所有型別的失業，不僅僅是由於人工智慧造成的。
我們的人工智慧驅動增長必須以公平、復原力和共享機會為基礎，這不會自然發生。維克拉姆·內爾先生問我們有什麼保障措施確保工人在人工智慧採納增加時得到公平對待。政府已制定了框架，如《智慧體人工智慧模型治理框架》和AI Verify，以為人工智慧供應鏈中的各方建立明確的責任，為人工智慧開發人員和使用者提供負責任實踐的清晰指導，包括人力資源技術解決方案提供商。薩克蒂安迪·蘇帕特先生恰當地指出，人工智慧採納在各行業、工人群體和不同規模的企業中以不同速度進行。如果不付出刻意的努力，人工智慧帶來的收益可能會流向某些人，而其他人被拋在後面。在中國，法院已經裁定純粹為了降低成本而用人工智慧替代員工是違法的。
國務部長陳振聲和三捷夫·庫馬爾·蒂瓦里先生談論了新加坡全國工會聯合會近年來在為工人配備人工智慧相關技能和支援勞動力轉型方面所做的工作。這正是我們應該利用的那種能力，以確保更多工人和企業瞭解可用的支援，以及人工智慧採納可以在整個經濟中加快。這就是為什麼我們全力支援新加坡全國工會聯合會成立三方就業委員會的提議。三方就業委員會將採取協調的三方方法，動員企業和工人在人工智慧時代實現公平和復原力增長。
正如何錦賢副教授所指出的，人工智慧應該增強工人，而不是替代他們。我們將利用新加坡國家僱主聯合會的商業顧問和新加坡全國工會聯合會的合作轉型中心幫助企業以驅動增長和增強工作角色的方式採納人工智慧，優先選擇增強人類能力而不是替代他們的技術。我們將利用三方夥伴與工人、工會和僱主的強大聯絡，推動跨行業和職業階段的廣泛人工智慧培訓，以確保在人工智慧重塑我們旅程時沒有工人被拋在後面。
在那些必須進行重組的技術或企業中，我們將與企業合作，幫助工人進行轉向、提升技能和重新技能培訓。
我們也將特別關注對人工智慧對初級工作的影響感到擔憂的學生和年輕工人。高等學府繼續增強其課程以跟上人工智慧的進步。所有高等學府從今年下半年開始，將為其校友提供選定的人工智慧相關課程，並提供重大折扣，期限為一年。
進入勞動力市場的畢業生也可以利用教育部的SkillsFuture帶薪實習計劃，該計劃將課堂培訓與企業的在職培訓相結合，以建立僱主重視的技能和經驗。
林瑞生副教授呼籲擴充套件青年學徒制路徑。我們同意。結構化學習必須輔以真實的工作場所經驗。我們將繼續與貨幣管理局和新加坡資訊通訊媒體發展局等行業領袖合作，支援高增長行業的學徒制，從我們的GRIT等專案經驗中學習，我們已準備好在必要時改進和擴充套件這些專案。
議長先生，讓我總結一下。新加坡曾經經歷過深刻的破壞，從亞洲金融危機到非典，再到新冠肺炎。每一次，每一場危機我們之所以能度過，不是因為政府有所有答案，而是因為工人、企業和政府肩並肩站在一起。這就是三方制的力量。
在許多國家，人工智慧成為一場拉鋸戰。一方是工人，另一方是企業。進展受到質疑，信任受損。新加坡不必走上這條路。我們共同努力，把整個經濟蛋糕做大，並確保利益廣泛共享。
對於想知道自己立場的工人，我們總會有你們的一席之地。你們的經驗、你們的判斷，比以往任何時候都更加重要，對於你們對國家的承諾、多年來、幾十年來的支援，我們深表感謝。非常感謝。[掌聲。]
對所有年輕的畢業生，你們的想法、你們的動力，比以往任何時候都更加重要。你們的熱情、你們的求知慾、那種聯絡、那種政務部長賈絲敏·勞剛才談到的求知慾，比以往任何時候都更加重要，我們支援你們。
對於我們所有的企業，如果你感到不確定，不知道從何開始，你不必獨自解決這個問題。我們將與你並肩同行。我們將幫助你轉型，幫助你競爭，這樣你就能為你的企業和員工創造更好的機會。
我們不會把工作的未來、我們工作者的生計、我們新加坡人的命運交由偶然決定。我們將塑造一個包容的、前瞻的、以實際行動為基礎的轉型。
新加坡人永遠不會是人工智慧驅動未來中的無助乘客，而是我們人工智慧之旅起航時的共同副駕駛。我們將以新加坡的方式向前邁進，政府、僱主和工會攜手合作，確保我們的人工智慧轉型創造良好的就業機會，為每一位新加坡工作者開闢通往更美好未來的清晰道路，因為每位工作者都很重要。基於此，我對這項議案表示支援。[掌聲。]
下午6時50分 議長先生：在議會中，我們也在採納和擁抱人工智慧，同時我們在這一旅程中為我們的工作人員做好準備。Low先生有澄清意見嗎？議長先生，請先動議豁免。
英文原文
SPRS Hansard · Fetched: 2026-06-09
[(proc text) Resumption of Debate on Question [5 May 2026], (proc text)]
[(proc text) That this House – (proc text)]
[(proc text) 1. Recognises the transformative power of new technologies, especially Artificial Intelligence (AI), to drive Singapore’s next phase of economic development; (proc text)]
[(proc text) 2. Emphasises that Singapore’s approach to AI-enabled growth must be anchored in fairness, resilience, and opportunity for all; (proc text)]
[(proc text) 3. Resolves to equip and support workers and enterprises to seize new opportunities and advance together; and (proc text)]
[(proc text) 4. Affirms that economic progress must remain inclusive, and that Singapore must not have jobless growth, because every worker matters. (proc text)]
[(proc text) Question again proposed. (proc text)]
12.32 pm Mr Speaker : Mr Mark Lee.
Mr Mark Lee (Nominated Member) : Mr Speaker, building on the earlier speech by my colleague National Trades Union Congress (NTUC) Secretary-General Ng made last night, I would like to focus on how this Motion is made real at the enterprise level.
Sir, no jobless growth cannot be achieved through worker support alone. It must also be designed into the way enterprises transform, the way jobs are redesigned and the way our system supports firms to move with confidence.
History shows that in periods of technological change, the winners are not those who try to protect existing business models but those who understand their underlying strengths well enough to redeploy them into new arenas. Fujifilm is one example. It did not survive the collapse of film by trying to sell more film. It built on capabilities in materials, optics and imaging to move into skincare, diagnostics and healthcare technology.
In sectors where Singapore is already strong – advanced manufacturing, logistics and connectivity, finance and healthcare – AI is not replacing the industry itself but changing how value is created within the industry. That is the mindset Singapore needs now. Our task is not to preserve every job exactly as it is. Our task is to help our enterprises and our workforce recognise their strengths, adapt them and move into the next phase of growth together.
I will divide my remarks into three parts: first, what AI is actually doing for businesses; second, its impact on jobs and workers; and third, how we can build a clearer enterprise front door and a wider bridge forward so that businesses and workers can cross the transition together.
Mr Speaker, businesses adopt AI where it delivers measurable outcomes. The evidence is already emerging. In customer service, generative AI improves productivity by around 15%, with larger gains among less experienced workers. Firms are also seeing double-digit gains in software engineering while in operations and supply chains, AI is improving forecasting, reducing waste and optimising inventory and logistics.
Second, AI does not simply improve jobs. It reconfigures work. Global evidence shows how AI adoption is increasingly concentrated among higher-skilled workers, with firms re-organising work around smaller, more experienced teams supported by AI. This creates a double risk: first, productivity gains may accrue more to those already ahead, widening inequality; second, it risks eroding the bottom of the career ladder.
If entry-level roles are reduced too quickly, the traditional pathway where workers build judgement through experience is disrupted. Firms may still require mid-level capability but fewer workers will have had the opportunity to develop it. That is a structural issue.
At the same time, new roles are emerging, integrating AI into workflows, validating outputs, redesigning jobs and translating domain knowledge into solutions. Companies like DBS and Mastercard are using AI to handle routine queries and personalise responses at scale, freeing up human agents for higher-value work. We are seeing the same among small and medium enterprises (SMEs). MTM Labo, a skincare company, for example, uses an AI tool called Hana that supports customer enquiries in multiple languages, allowing its team to focus on more complex, high-touch interactions.
This is an important point. AI, when deployed thoughtfully, does not just replace jobs. It changes the nature of jobs and can raise the value of human work.
But adoption is not plug-and-play. It requires integration into workflows, redesign of processes and alignment with business strategy. This is where many firms, especially SMEs, face challenges in translating AI into implementation. If we do not address this, capability will concentrate among larger firms and among more skilled workers. The gap will widen. That outcome would run directly against the spirit of this Motion.
However, we must also be careful about how we respond to these changes.
I understand Mr Ng's good intent behind the call for earlier notification of retrenchment to better support workers. But if firms are not yet ready to redesign jobs or absorb workers differently, earlier notification alone will not solve the problem.
If AI is re-organising work, perhaps a better solution that matters more is not whether we intervene earlier after displacement occurs but intervening early enough before displacement becomes necessary. We should shift the focus from managing retrenchment to enabling all firms to redesign jobs and retrain workers so that they transform and the workforce adjustment happens alongside transformation, not after it.
This brings me to my third point – how do we we build a bridge forward? A bridge that is wide enough for many to cross, not just a select few.
The recently introduced Enterprise Workforce Transformation Package is a step in the right direction. With the Singapore Business Federation (SBF) and the Singapore National Employers Federation (SNEF) as programme partners, companies can access advisory support to redesign processes and job roles and tap on the SkillsFuture Workforce Development Grant for consultancy, workforce technology adoption and capability building. This is a meaningful shift.
However, today, firms are still navigating multiple schemes, multiple agencies and multiple claims processes, slowing adoption at the point where speed matters most. But I am heartened to hear from Minister Tan last night that the merger of SkillsFuture Singapore and Workforce Singapore into Skills and Workforce Development Agency aims to solve this issue.
But can we do more to help businesses?
This brings to my first dimension of practicality. One way forward is also for AI grants and schemes to have a more integrated approach where businesses can access this support through a single interface rather than navigating multiple agencies and deploy it flexibly, whether for basic implementation, subscriptions or experimentation, without repeated layers of claims, just like the SkillsFuture Enterprise Credit wallet.
For larger or more complex customised projects, there may also be a case for more upfront support so that firms are not constrained by cashflow when making longer-term investments.
At the same time, we must recognise that AI adoption involves experimentation. Not every project will succeed. If firms are penalised despite genuine effort, we risk discouraging innovation. A reasonable tolerance for failure will be necessary if we want companies to move decisively.
The second dimension is capability at scale. The pool of experienced AI professionals remains small and highly competitive. If we are too restrictive on bringing in foreign AI talents, we will slow down capability building across our economy. So, our efforts to build a wide bridge must also mean keeping our talent pipelines open. Not just to bring in talent, but to allow that talent to cross-pollinate skills across firms and sectors.
For SMEs, one area we can consider is to give targeted flexibility to bring in specialised AI expertise beyond existing manpower constraints. This can be time limited and on an application basis, with some checks in place to prevent abuse.
The third dimension is through our institutes of higher learning (IHLs). IHLs can be positioned more deliberately as execution platforms for applied AI, especially through Centres of Excellence anchored around postgraduate programmes. Many postgraduate students, including foreign talent, bring prior industry experience and technical depth. When paired with local undergraduates, this creates a practical model for capability transfer. If anchored around real SME and sectoral problem statements, these teams can go beyond proof-of-concept work to develop deployable solutions.
This achieves several outcomes. It lowers the cost of experimentation for SMEs while enabling cross-pollination between international and local talent, and it creates a pipeline for startups to emerge, anchored in Singapore and focused on solving real industry demands.
This model also helps address to the broken career ladder. As entry-level pathways narrow, embedding students in real problem-solving builds capability earlier. At the same time, it strengthens our students' critical thinking, preparing them to question and validate AI and not just rely on it.
The fourth dimension is industry enablement. Today, many firms are attempting to solve similar AI problems in isolation. This leads to duplication of effort, higher experimentation costs and slower adoption.
Trade associations and chambers are structurally positioned to address this gap. They operate at the interface between government policy and firm-level behaviour and can translate national AI strategies into sector-specific implementation. They can function as coordinating platforms, identifying common industry problem statements, aggregating demand and working with solution providers and IHLs to develop integrated, deployable solutions aligned to actual workflows and job roles.
Let me illustrate. SBF is developing an AI tool to help businesses understand the rules of origin of free trade agreements and how to apply preferential tariff treatment when goods are exported overseas. This is one example of how an industry‑led approach can reduce duplication and improve efficiency.
For many SMEs, the challenge is also on applied capability. Firms want to know which problems to prioritise, who can help and how to proceed without excessive cost or risk.
The Singapore Chinese Chamber of Commerce and Industry (SCCCI) AI Experience Programme, launched together with the Infocomm Media Development Authority (IMDA) in support of the Digital Enterprise Blueprint shows this clearly. It has been heavily oversubscribed because SMEs are looking for guided entry points.
Similarly, SCCCI's AI Enablement Programme allows SMEs to define real problems and work with students from Nanyang Polytechnic, Singapore Polytechnic, Temasek Polytechnic and the Singapore University of Technology and Design (SUTD) to develop solutions. SMEs gain workable solutions. Students gain relevant experience. And knowledge spreads across the system. With the right support and funding, trade association and chambers can become platforms that accelerate AI adoption across sectors.
Finally, Mr Speaker, the fifth and most important dimension is human capital. The success of AI-enabled growth will not be determined by how many tools we deploy, but also by how many workers we carry through that change.
I agree with Mr Ng that workers who are displaced, whether due to AI or industry consolidation, need stronger support during transition. But we should also consider how that support is structured.
Today, much of it follows a "train and place" approach, where workers are first retrained and then supported to find a job. In practice, this can be uncertain. No worker wants to be retrenched, spend months in training and still face uncertainty about the next job. We should, therefore, move more deliberately towards a "place and train" model. If another company is prepared to take in a displaced worker, even if the fit is not immediate, we should support that transition directly. This can be done by targeting temporary wage support to the receiving employer, similar to the spirit of Jobs Support Scheme (JSS) but anchored under an industry transition fund so that companies are incentivised to hire first and retrain on the job.
This shortens the period of uncertainty for workers, while giving the firms the confidence to take in and develop new talent. This is also where trade associations and chambers, and unions can also play a coordinating role – identifying firms with demand and matching them with workers at risk.
Finally, Mr Speaker, the open bridge must be a moral and social contract, and at the heart of that contract is trust. Workers must see AI as enabling, not threatening. If AI is seen as a tool to remove jobs or close off pathways, adoption will slow not because the firms lack technology but because trust is lacking.
And if that trust is broken, we may inadvertently create a lose-lose outcome where governments step in with more restricted workforce regulations around AI adoption, raising longer-term costs and reducing flexibility for businesses. Trust must, therefore, be built deliberately through how AI is deployed, how jobs are redesigned and how workers are supported through change.
Sir, I am a businessman and I join the call to support my union brother Mr Ng in this Motion because I believe this embodies the true spirit of tripartism that has served Singapore well.
We cannot say every worker matters and then leave workers to navigate this transition alone. Businesses must lead in redesigning jobs and investing in their people. Workers must step forward and adapt. And Government must ensure the system enables both. Only then can workers and businesses advance together.
Mr Speaker, this motion reflects Singapore's determination to get this right. Let us take the proactive path to work together, to build trust early and ensure that AI expands opportunity rather than narrows it. Sir, I strongly support the Motion. [ Applause. ]
Mr Speaker : Mr Saktiandi Supaat.
12.49 pm Mr Saktiandi Supaat (Bishan-Toa Payoh) : Mr Speaker, before I begin, I would like to declare that I am working in a bank, a financial institution in Singapore. I am also an advisor to the Union of Power and Gas Employees (UPAGE) and the Logistics and Supply Chain Union (SCEU).
Mr Speaker, I would like to thank, foremost, the Secretary-General of NTUC and Member, Mr Ng Chee Meng, for moving this important Motion and for setting out the need for a new compact for AI-enabled growth, one that keeps workers at the centre of our transformation, anchored in fairness, resilience and opportunity for all.
I will focus on how we can build a more inclusive AI economy, where growth is not only strong, but broadly shared.
AI is no longer emerging; it is already reshaping how we work and live. Beyond well-known AI tools, I saw this first-hand during engagements with the UPAGE, SCEU and, recently, Manpower Government Parliamentary Committee's learning journeys to SMRT, NTUC Finest at Punggol and Chye Thiam Maintenance Pte Ltd.
In different sectors, power and gas, supply chain, transport, retail, food and beverage (F&B), cleaning and facilities management, all of them embedding AI, autonomous vehicles (AVs) and robotics into workflows, raising productivity, creating new jobs and redesigning roles for existing workers.
There is now a global race in AI innovation and adoption. And there is a growing belief that those economies and companies that move early and decisively will capture the greatest value.
But Mr Speaker, Sir, as this House has consistently emphasised, economic success must be inclusive. Economic success is not just growth, but how widely its benefits are shared. Uneven growth is not a model Singapore should follow. Besides the redistribution measures that we have implemented through tax measures and targeted assistance, we must cultivate a sustained mindset to ensure AI-enabled growth is inclusive.
Usage of AI has raised concerns for many workers. For example, will AI take away my job? Will it lead to jobless growth?
Based on the NTUC's survey on economic sentiments in Singapore, which it conducts yearly, the fear that AI would replace their job or current role is more pronounced for professionals, managers and technicians (PMETs) and entry-level jobseekers. For non-PMEs and lower-wage workers, the lower level of concern could be because they do not use AI tools as extensively and are unaware of how AI would impact their job opportunities and not because there is no AI-disruption risk.
These concerns are reinforced by news of job restructuring globally, as well as uneven adoption of AI across sectors and occupations.
Mr Speaker, Sir, let me illustrate this unevenness with a concrete example from our own financial sector, of which I am working in. In banking here, AI is no longer emerging, it is already deeply embedded and a core driver of productivity. Across the sector, our local banks are deploying automation in operations and customer servicing, particularly in labour-intensive and repeatable processes. These are not experimental use cases; they are transforming how work is done across the value chain.
AI is now used in operations, such as processing, settlements and compliance, and in credit evaluation through machine learning models that support faster and more consistent underwriting. It is even beginning — well, I would not say beginning; it is already playing a bigger role in hiring, particularly in initial screening.
But the impact is not uniform. Routine clerical and processing roles are most exposed, while hiring is shifting towards higher-skilled roles in data, AI, cybersecurity and governance. At the same time, regulation is sustaining demand for oversight roles in risk, compliance and audit.
But crucially, more complex work still requires human judgement. Handling nuanced customer issues, managing relationships and making judgement calls cannot be easily automated. As a result, many roles are not disappearing but evolving. For example, credit officers are moving towards interpreting and oversight while contact centre roles are shifting towards experiential, escalation and trust-building. So, what we are seeing is not wholesale job replacement, but a reconfiguration of tasks within jobs.
Mr Speaker, Sir, while much of the AI discussion focuses on PMETs, we must not overlook skilled tradespersons and our blue-collared workers. Our electricians, technicians and maintenance workers are essential. AI cannot repair lifts or maintain MRT systems on its own. As our economy becomes more digital, these roles will become more sophisticated, not less. They are, in fact, high-mastery professions.
If AI raises the premium on skills, we must also raise how we recognise mastery. But beyond recognition, we must also rethink how mastery is built. As AI reshapes how work is performed, we must also rethink how skills are transmitted.
Today, many of our industry transformation maps (ITMs) guide sectoral growth and workforce development. But they were largely designed for a pre-AI world. There may be merit in updating these frameworks to explicitly account for how AI is changing apprenticeship and on-the-job training pathways. This issue has been raised in this House before and it deserves renewed attention.
In particular, we should consider whether we need Industry Training Continuity Maps alongside our ITMs, to ensure that even as AI takes over more routine tasks, we continue to sustain a strong pipeline of deeply skilled human workers especially in roles where mastery, judgement and hands-on expertise cannot be replaced.
Today, this is less visible in skilled trades. This is why I have also proposed a National Master Trades Accreditation framework during the last Committee of Supply, to recognise progression, reward deep skills and integrate AI competencies.
If we get this right, AI will not hollow out middle-skilled jobs, it will elevate them. This will also help elevate further the skilled graduates from our Institutes of Technical Education (ITEs) and polytechnics.
Mr Speaker, Sir, given the immense opportunities and risks, AI as a growth engine requires sound policies that benefit workers and citizens. Countries, such as the United Arab Emirates and Finland, have adopted coordinated national AI strategies combining government leadership, enterprise adoption and workforce development.
Here in Singapore, we have taken important steps. Budget 2026 announced the establishment of a National AI Council led by Prime Minister Lawrence Wong, alongside incentives for businesses through the Enterprise Innovation Scheme and support for workers via SkillsFuture and the TechSkills Accelerator.
I would like to offer some suggestions to equip workers and enterprises more effectively, so that no worker is left behind.
First, we must normalise AI as a way of life. We must go beyond training pathways and time-limited access to AI tools. While I welcome the provision of temporary access through Government and NTUC initiatives, we must think ahead about what happens after this initial phase of subscriptions.
Many of the more capable AI tools require ongoing subscriptions. Over time, this may create a divide between those who can afford to use these tools regularly and those who cannot. If left unaddressed, we risk creating a new form of inequality between AI "haves" and "have-nots".
Because access to AI tools directly affects productivity, learning and income potential, unequal access will translate into unequal outcomes. We should therefore consider how to ensure sustained and affordable access, especially for lower-income workers, freelancers, tradespersons and small businesses.
Possible approaches include a baseline level of subsidised access, similar to digital connectivity, tiered or group-based pricing with industry partners, shared access through community centres, libraries and training hubs, and ensuring employers receiving AI support also extend access to workers.
Mr Speaker, Sir, if AI is to be a force for inclusive growth, access cannot be a privilege, it must be broadly shared. And in the digital age, access to AI may become as fundamental as access to the Internet. We must ensure no Singaporean is priced out of that future.
One way to drive adoption is for the Government to be the "first customer" of useful AI tools and to mitigate AI transition-related concerns. AI-enabled systems can provide faster and more practical responses to citizens navigating Government services. As AI systems consistently deliver useful outcomes and quick advice to Singaporeans, confidence in AI and AI adoption will grow.
Against this backdrop, I would also like to acknowledge the Government's efforts to support workers through an AI-shaped economy, with employment remaining the central outcome and address job anxieties.
As such, it is important to recognise, for example, that employment outcomes for the Malay/Muslim community under M³ and Focus Area 4 (FA4) have been delivered at both scale and through targeted support. Between 2022 and 2025, Workforce Singapore and NTUC's Employment and Employability Institute (e2i) assisted over 29,000 Malay/Muslim jobseekers, with more than 19,000 placed into jobs.
At the same time, through community-based pathways under M³ and FA4, over 6,000 jobseekers were engaged, with more than 500 securing employment, including those requiring more sustained support. This reflects both the breadth of our national employment system and the depth of our community-anchored interventions. Building on this, FA4 workstreams will sharpen their focus on supporting workers through an AI-shaped economy, with employment remaining the central outcome.
I am glad that NTUC will work closely with MENDAKI through NTUC's e2i to strengthen job transitions, particularly for young adults transiting from campus to career, who may be entering the workforce amid heightened uncertainty about job relevance and AI-driven displacement. This includes strengthening early career pathways by partnering IHLs and integrating engagements with e2i's career services, job matching and employer networks, so that fresh graduates are better prepared for a changing labour market.
For underserved Malay/Muslim workers, as an early step, MENDAKI and NTUC's e2i jointly piloted Langkah Digital AI workshops in community settings, with plans to scale further this year; and I plan to attend some of these workshops. Taken together, these deliberate employment-linked interventions will help to ensure that productivity gains from AI do not result in jobless growth, but instead, equip Singaporeans, across life stages, to adapt, remain employable and progress with confidence, supported by NTUC, its partners and the wider Labour Movement. So, I encourage our community to take advantage of these initiatives to upskill.
Second, we must focus on the infrastructure that enables AI. Singapore has invested in strong digital rails. Systems, such as Singpass, already allow secure transactions, including legally binding processes, such as the Lasting Power of Attorney. The next step is to enhance interoperability through application programming interfaces (APIs), so more services can be integrated seamlessly.
When services are integrated, AI can significantly enhance efficiency and user experience. At the same time, we must calibrate data-sharing frameworks and safe harbours, so that data can be used responsibly without stifling innovation.
Third, we must ensure employers redesign workflows to embed AI meaningfully. Member Mark Lee has mentioned that. Training alone is not sufficient. While the Enterprise Workforce Transformation Package provides useful support, it tends to reach large firms that are already inclined to transform. We need to go further.
One possibility is to launch an "AI Bilingual" accreditation, for employers and not just for workers and jobseekers. While it can be on a voluntary "opt in" basis, like BCA's Green Mark Certification Scheme or TAFEP's Fair Employment Badge, the accreditation can be tied to certain other benefits or quotas to incentivise companies to come forward. Like existing voluntary schemes, this could be linked to incentives to encourage broader participation.
Fourth, we must support those with traditional employers, tradesmen, lower-wage workers and platform workers. AI can act as a personal assistant, enhancing productivity and income. For example, tradesmen can use AI tools to generate quotations, invoices and customer responses; lower-wage workers can use AI for scheduling and financial planning; and platform workers can optimise routes and jobs across platforms, increasing autonomy. Is there scope for the Government to invest in such tools and provide time-limited access, so these workers can experience their practical benefits?
Finally, we must recognise that not all workers have equal capacity to adapt to AI. Time constraints, caregiving responsibilities and life-stage challenges affect participation in training. We should move forward towards flexible learning; integrate training into work; and strengthen cross-sector mobility. This ensures our workforce remains agile, mobile and inclusive. Mr Speaker, Sir, allow me to now to speak in Malay, please.
( In Malay ) : [ Please refer to Vernacular Speech .] We want every worker to receive the support they need to upskill and not be left behind as AI continues to progress. Through my interactions with Malay/Muslim workers, many see AI as an opportunity, but also worry that they cannot keep pace with this rapid technological advancement.
This concern is valid. AI is transforming the way we work. PMETs in particular are beginning to ask: are my skills still relevant? Can I adapt to these changes? The fundamental question is: Does this AI economy have a place for me, or will I be left behind?
As co-chair of the Economic Resilience Committee alongside Dr Wan Rizal, our focus is clear – to build and strengthen the economic resilience of our community by embracing the AI transition with openness and readiness. We want growth that creates opportunities and empowers workers, not replaces them.
We will assess the implications of the economic shifts that the Economic Strategy Review Committee will outline, identify new sectors and growth opportunities and understand how we can encourage broader participation from the Malay/Muslim community in these areas. At the same time, we are developing targeted strategies to strengthen community involvement in economic transformation initiatives, so that participation can be deepened across all segments – from youth to professionals and entrepreneurs.
Within the Malay/Muslim community, this work has already begun and must be strengthened through the M³, now known as M³+. Allow me to share some examples of efforts being undertaken by Malay/Muslim institutions to raise AI literacy among our community.
We are seeing encouraging initiatives. At Majlis Ugama Islam Singapura (MUIS) (or Islamic Religious Council of Singapore) and the mosques, AI literacy programmes help the community understand the responsible use of technology. The IftaSG initiative uses AI for Fatwa research. Programmes encouraging thoughtful AI integration have also been made available to asatizah, to enable richer and more meaningful Islamic learning experiences.
At Persatuan Ulama dan Guru-Guru Agama Islam Singapura (PERGAS), AI training equips asatizah with digital skills through programmes such as Diversity-driven Upskilling for Asatizah, the AI Accelerator Challenge, and AI for Asatizah Entrepreneurs.
At the Association of Muslim Professionals, for working professionals, the "Learning Circles: All About GenAI" programme will be held this month. It will explore how generative AI is reshaping the way we work and how professionals can begin applying it meaningfully in their daily roles.
At MENDAKI, the MENDAKI Achievement Programme, or MAP, now uses AI tools such as Khanmigo and KiteSense Luminee to enhance the learning outcomes of students from less privileged backgrounds. This reflects a commitment to ensuring that technological progress serves as a catalyst for social mobility and inclusive growth within our community.
Mr Abdul Kadir bin Abdul Rahman, a veteran educator in science and mathematics, is a fine example of how three decades of deep experience can be combined with present-day innovation. Now a trainer in the MAP programme, he is a strong advocate for using AI technology to improve learning quality. He has noted that one significant change is students' increased willingness to ask questions – fostering a more interactive and supportive learning environment.
In addition, MENDAKI's Langkah Digital initiative provides AI-Ready workshops, hands-on training and upskilling programmes to help individuals understand and apply AI in their lives and work. MENDAKI has partnered with institutions, such as SUTD, opening up opportunities in AI, design and application-based learning – so that our community is not merely a consumer of technology, but capable of mastering it.
This shows that AI can be a catalyst for social mobility – if we ensure that access and opportunity are widely shared. But not everyone starts from the same point. Some have access, a supportive environment and the time to learn. Others face constraints – whether in terms of time, family responsibilities, or self-confidence. That is why our approach must be inclusive. Training must be accessible, relevant and practical, so that every individual has the opportunity to adapt and progress alongside these changes. Ultimately, success in the AI economy is not measured by technology alone, but by how well we ensure that every citizen can move forward with confidence and hope.
( In English ): Mr Speaker, Sir, AI will bring both disruption and opportunity. If managed well, it can raise productivity and expand opportunities for all. But if left unmanaged, it can widen inequality.
We must ensure that AI drives not just growth, but inclusive growth. If we get this right, AI will not divide our workforce, it will strengthen it.
And in doing so, we will renew this compact for AI-enabled growth. One where every Singaporean, whether working with code or with their hands, has a place and a role and a future in our economy. I wholeheartedly support this Motion, Mr Speaker.
Mr Speaker : Ms Yeo Wan Ling.
1.09 pm Ms Yeo Wan Ling (Punggol) : Mr Speaker, not long ago, I was at a stop light in Punggol when an autonomous shuttle glided past. I was not the only one watching. Around me, drivers and pedestrians looked up – a mix of curiosity and something quieter. A low, humming anxiety, fringed with a dash of awe. And behind their eyes, a very human question: what does this mean for me?
That image has stayed with me. AI and autonomous technology are transforming the way we work, live and play, faster than any technology humankind has seen. So, the question this House must answer today is not how these technologies work, but what we are doing – concretely, deliberately – to make sure our workers' lives and livelihoods are not left behind.
Mr Speaker, that is what this Motion is really about. And I want to speak to it not with mere assurances, but with a plan. A plan for our workers, a plan for our union members, a plan for our brothers and sisters sitting up in the gallery supporting us in this Motion.
The transformation is already happening, quietly, all around us. At Changi Airport, autonomous baggage tractors ferry luggage between terminals. At Marina Barrage Service Road, autonomous sweepers clear leaves and litter. At Pasir Panjang Terminal, driverless automated guided vehicles move containers between yards. And our first revenue-generating autonomous bus services are set to run on two routes in the second half of this year.
Upskilling has rightly been at the forefront. But job redesign is equally critical to re-engineer existing jobs for new realities, to create new job types and to support workers through transitions as AI reshapes job longevity. And for job redesign to truly move the needle, it must be a genuine, ground-up effort with workers and their real workflows at the centre. Let me elaborate.
First, consult workers deliberately, to really understand their work. As Executive Secretary of the National Transport Workers' Union (NTWU), I have seen first-hand what it looks like when tripartism works. Our management partners, SBS Transit, SMRT and others, have been preparing our bus captains and technicians for the advent of AI, electric vehicles (EVs) and AVs. Employers provide upskilling and training. Government supports companies and workers through that process. And unions, we do what we do best at: listening carefully on the ground to what workers truly need.
It is exactly that ground listening that surfaced something we would have otherwise missed. Even as we move towards 50% of our bus fleet becoming electric by 2030, our bus captains flagged that EV training had important gaps. Unlike conventional buses that use mirrors, EVs use digital monitors and our captains told us about the time delay, the glare, the eye strain and in more severe cases, nausea. They asked for longer training and preparatory times. The union pushed for it with our tripartite partners and it was addressed.
Mr Speaker, that is feedback no consultant's report would have surfaced. But it directly shapes bus design, driver safety and passenger experience. Workers know their jobs better than anyone. That is a resource we must keep on tapping.
In anticipation of our first AV bus services, NTWU, last year, surveyed around 500 bus captains and technicians. One in three expressed concern that AVs would affect their jobs – job security was the top worry, followed by fears of pay cuts. Unsurprising. These are sentiments shared by transport workers worldwide. Yet one in three of our Singaporeans, also remained confident that drivers would continue to play an important role.
So, we dug deeper. We sat with bus captains and asked them to walk us through a day's work, not what their job description said, but what they actually did.
What they told us turned our assumptions upside down. On paper, we assumed driving was the core of a bus captain's job, perhaps, 80% of their tasks. Our captains told us it is closer to 20%. The other 80%, helping elderly passengers board safely, managing crowding, de-escalating difficult situations, giving directions, being a calm and reassuring presence on board and even telling passengers they can only have a singing performance when seated – these are deeply human responsibilities that no AV can replace.
This has profound implications. If we had acted on our paper assumptions about the bus captain role, we would have misjudged job sizes, skill requirements and salary structures – creating inequitable outcomes for workers and human resource (HR) planning disasters for organisations alike. Getting the job description right is not a bureaucratic exercise. It is the foundation on which all of job redesign rests.
Mr Speaker, the unions and the Tripartite Jobs Council will continue to walk the ground. But we cannot do this alone, not if we are serious about job redesign at scale. I call on the Government to resource this properly: to fund systematic study and mapping of actual job roles and workflows, so that job redesign is built on ground truths and not just assumptions.
Mr Speaker, my second point is this: for AI transformation to succeed, workers and customers must be at the centre of the reimagination process of what AI can bring to workers and businesses. Not consulted after the fact. Not informed of decisions already made. At the centre, right from the start.
AI will transform work as we know it. But where exactly it will land – which tasks, which roles, which industries – nobody can fully predict that. That is precisely why the reimagination process matters so much. We cannot wait until the dust settles. We have to build, prepare and yes, dare to dream of what an AI-powered workplace can look like together with our workers. The most important lesson I have taken from visiting companies in transformation is this: when you involve workers early and genuinely, they do not resist change. They drive it.
Mr Speaker, let me tell you about Trusted Hub. A Singapore SME, 25 years in business, in the business of data processing, which, at its heart, actually is what AI is. Rewind to 2001, Trusted Hub was handling Government submissions from members of the public; stacks and stacks of paper; photocopiers, faxes, prints. Fast forward to 2026, same business, more or less the same clients, but a completely different way of working. AI now processes much of the data, taking the load off their staff.
What impressed me when I visited was not the technology. It was the people. Because Trusted Hub brought their workers into the reimagination process as stakeholders – not passengers – the majority of their staff have upskilled themselves to programme AI Agents, creating both enterprise and innovation value for the company. And the oldest AI Agent programmer in the company? A gentleman in his 60s. Self-taught. And this is what happens when you do not underestimate your workers.
I have spoken in this House before about FairPrice's Store of Tomorrow at Punggol Coast Mall, featured at international trade shows as a model for the supermarket of the future and a living showcase of how technology can make work better, easier and safer for our workers. It displaces fear not with words but with evidence people can walk into and see for themselves.
But what made it work was not the AI. It was the process. Workers and unions shaped and designed the system, not inherited it. And because of that, staff did not just accept the change, they owned it.
And what I want is more of this. Stores of tomorrow, bus interchanges of tomorrow, restaurants of tomorrow, clinics of tomorrow. Living testbeds that allow reimagination to happen, not just within companies but across clusters and our communities, so that conversations about AI in the workplace can take place openly, candidly and with imagination, rather than dread.
Very much like my Punggol residents watching our autonomous shuttles glide by, a low hum of anxiety, yes, but definitely fringed with a dash of awe.
While the shape of tomorrow's workplace is still forming, one thing is clear. Putting workers and work processes at the centre of transformation is not optional, it is the method. What does it look like in practice?
It is Chye Thiam Maintenance offering $200 training allowance to workers who volunteer to be trained on their robo-sweepers, making transformation something workers choose and not something done to them.
It is Grab working with unions to assess whether an AV Shuttle safety driver can sustain a full eight-hour shift on continuous alert because worker welfare is part of the design and not an afterthought.
And it is a British entrepreneur who has started calling his AI bots, AI employees, to remind himself and his team that AI is not about replacing people but about changing roles.
These are not grand gestures. They are small, deliberate, but very significant acts that normalise AI in the workplace and make it something workers can see themselves thriving in rather than being displaced by. This involves responsible employers, progressive employees and indeed, a supportive and nurturing Government. It is the tripartite way and it is why the Tripartite Jobs Council matters so much to organise, to set the tone right from the start, on how AI is embedded and rewarded in everyday company life.
Mr Speaker, this is the real answer to unfounded fears about AI displacement – not reassurances, but evidence. Evidence that when workers are treated as co-creators, transformation is faster, adoption is stronger and outcomes are better for everyone.
The Tripartite Jobs Council is well-placed to drive this reimagination work at every level, through Company Training Committees (CTCs), CTC Queen Bees at cluster level , sectoral AI uplift plans across industries. Our Queen Bees can bring their contractor ecosystems along, as what FairPrice did with their Store of Tomorrow. Unions will do what we do best: walk with workers and management through what the road ahead looks like, and what new roles are emerging along the way.
But Mr Speaker, this really requires investment and intentionality. I call on the Government to resource sectoral AI uplift plans for industries – retail, logistics, healthcare – with the same deliberateness that has started to guide our AV roadmap for public transport. Workers deserve to know not just that AI is coming, but where the next testbeds will be, what the new jobs will look like and how to get there. Clarity is not a luxury. For workers standing at that crossroads, it is everything.
Mr Speaker, my third point is this: even the best-run AI transition will see jobs disappearing and some occupations finding the tasks they do taken over by AI. That is the honest truth. And we should not paper over it with optimism. Hence, transition support must be real, it must be timely and it must reach those who need it most.
We owe our workers a system that catches them before they fall too far and gets them back to a good job as quickly as possible. That system must start with job redesign. Not as an afterthought but as the first line of defence. If we redesign jobs well and early, we reduce the number of workers who need to be caught in the first place. The best transition support is one that makes the cliff shorter to begin with.
That is why the signal we send to enterprises matters so much. AI grants must be tied to mandatory job redesign requirements and productivity gains linked to worker outcomes. If these enterprises are unable to retain our workers, these companies should be required to notify the Government early on personnel whom they are unable to retain, so that these displaced workers can be assisted by e2i and our newly-formed Tripartite Jobs Council. This will be the assurance to workers that Singapore's AI transition will not result in jobless growth and that we keep the transition time to a new good job as short as possible.
Mr Speaker, I am heartened by our Prime Minister's assurance during Budget 2026 that the AV transition will be managed carefully, with close engagement with Platform Worker Associations and our drivers. As Advisor to the National Taxi Association and the National Private Hire Vehicles Association, I want to speak directly to that. In Mandarin, please, Speaker.
( In Mandarin ) : [ Please refer to Vernacular Speech .] Our taxi and private-hire drivers are already navigating a fiercely competitive environment with high fuel costs. Seeing AVs operating in Punggol and hearing news of autonomous bus pilots, they cannot help but harbour a quiet, unspoken worry. They are not asking us to halt technological progress – but what they need is not just reassurance. They need a clear sense of direction.
How will the geofencing of autonomous vehicles be progressively expanded? What is the timeline? New roles such as remote operators and safety supervisors are beginning to emerge. I hope that drivers who are willing will be supported and given access to training, so that they can transition into these new positions. For those who are not yet able to make that transition, I also hope that the newly established Skills and Workforce Development Agency will take the time to understand their needs more carefully – because they are not a homogeneous group and cannot be treated as one.
This is not just a matter that concerns platform drivers. Our technicians and tradesmen keep Singapore running with their hands, yet in conversations about AI, they are often invisible. Their contributions have long gone without sufficient recognition.
I am glad that the Ministry of Manpower (MOM) has begun driving efforts in this area, starting with the electrical trade. We must press on to build career pathways for tradesmen that are more promising and more respected, while harnessing AI to enhance their capabilities – not to replace their judgement.
( In English ): Mr Speaker, the workers most exposed to AI disruption are often the ones with the least buffer – least savings, least flexibility, least time to wait for a system to catch up with them. That is why our response must be tripartite in the fullest sense. Employers and platform partners must lean in as their business models evolve and not step back. This means staying involved in transition support, co-sharing training cost, covering opportunity cost and supporting workers through employment and post-employment pathways.
Unions will do what we have always done, walk the ground, listen and shape livelihood opportunities alongside our workers and we will continue to "jaga rumah" – by keeping watch on what other jurisdictions are doing, from China's Internet court ruling that AI replacement alone is not grounds for dismissals to California's requirement for human safety operators on AVs. These are signals of a world working out where the boundaries are. Singapore must learn from them, and when necessary, get ahead of them. Government must design transition support around the people who need it and not around what is administratively convenient. This is a standard we must hold ourselves to.
Mr Speaker, I will conclude with three calls to the Government.
Give time. Job redesign cannot be rushed. It requires sitting with workers, understanding what their work really is, not just what the job description says, and going back to the ground when the first answer turns out to be incomplete, as ours was with bus captains. Companies need to be supported through this and not just pushed into it.
Give help on the reimagination piece. Most companies, especially our SMEs, cannot do this alone. I look to the Government to provide practical facilitation, frameworks and funding that makes job redesign achievable. And I look to our leading companies, our Queen Bees, to step forward, share what has been worked on, and bring their sectors along with them. Transformation that stays within one company is transformation that is only half done.
Make NTUC the linkway. Our unions, our e2i, our Tripartite Jobs Council – we are already on the ground, in the companies, with our CTCs, sitting across the table from workers and employers every day. We have trust that took many, many years to build. Our Labour Movement is ready to be the connective tissue of this transition, matching displaced workers to redesigned roles, advocating for fair treatment and holding everyone, including ourselves, to account. The SWDA and our agencies must build on this.
Mr Speaker, I think back to that autonomous shuttle gliding through Punggol. Our workers watching it are not asking us to stop it. They are asking us to make sure that as it moves forward, they move forward too. That is the answer we owe them. Not just a promise; a plan.
I believe that an AI transition with no jobless growth is possible. Not because the technology will take care of it, but because we will. If we consult workers properly, involve them in the reimagination of their work and back that up with transition support that actually reaches the people who need it most. I support the Motion.
Mr Speaker : Mr Gerald Giam.
1.28 pm Mr Gerald Giam Yean Song (Aljunied) : I declare my interest as the owner and director of a company that provides software to training providers.
Mr Speaker, we face a structural threat to our workforce. For decades, Singapore's economic model has been built on the premise that a highly educated and skilled workforce would hold the keys to a prosperous future and be a buffer against economic storms. However, we are now in the midst of a paradigm shift where AI is not only augmenting human capability, but in many ways replacing it. Unlike past economic cycles, where such turbulence could be written off as an episode of creative destruction, AI promises to be a harbinger of a fundamental shift in our economic and social relationships. Taking this concept further, it would even impact the roles that the Government plays in mediating between the individual and society.
Today, we must recognise that the very nature of labour's economic power is changing. Failure to address this issue, even as productivity soars, will lead to an entrenched lower and middle class with the loss of economic agency. This concern is articulated by Jasmine Sun in an opinion piece for the New York Times, where she identifies the San Francisco consensus, a growing recognition that the hiring of young workers in highly AI exposed occupations is already in decline. She reminds us of the risk of a resulting permanent underclass, where the gains of technology are concentrated in the hands of a very few.
Not all the evidence points towards catastrophe. A 2025 US National Bureau of Economic Research working paper found that tasks with higher AI exposure do experience reduced labour demand. However, overall employment effects have so far been modest as productivity gains offsets some displacement. Similarly, a study published in the Quarterly Journal of Economics by the Massachusetts Institute of Technology's Danielle Li and Stanford's Erik Brynjolfsson found that generated AI tools boosted worker productivity by nearly 15%, with the greatest gains among the less experienced workers, thus suggesting AI can be a ladder, not just a trapdoor.
It should be noted that these studies examined early and controlled deployments. As agentic AI scales across entire industries simultaneously, the distributional consequences may be more severe and swifter than early productivity research would suggest.
We cannot be certain which trajectory Singapore is on. The asymmetry of risk demands that we prepare for the harder scenario, not the easier one.
This concern is shared by the very architects of the AI revolution. In 2021, OpenAI's CEO Sam Altman predicted in his blog post "Moore's Law for Everything" that AI would shift power from labour to capital, positing that if public policy does not adapt accordingly, most people will end up worse off than they are today. Crucially, Altman was not fatalistic. He argued that the proactive redistribution of AI-driven wealth, including giving citizens equity stakes in the economy, could make this a broadly prosperous transition.
Similarly, Dario Amodei, the CEO of Anthropic, has observed that the health of a democracy is premised on the average person having leveraged through creating economic value, a view he expressed in his 2024 essay, "Machines of Loving Grace".
The erosion of that leverage is a deeply concerning prospect that requires a bold and structural policy response. Singapore is uniquely positioned to lead this response and to capture the genuine economic opportunities AI presents for our people. As a small, open economy with a highly educated workforce, strong institutions and well-capitalised sovereign wealth funds, we have the tools to act swiftly and structurally, compared to many large nations.
But that window of opportunity will not remain open indefinitely. While cost arbitrage makes offshoring attractive, AI could erode that advantage – not by bringing those jobs back but by enabling small teams of skilled Singaporeans to do the work that once required hundreds of offshore workers. The opportunity is not in reshoring in a traditional sense but the concentration of higher-value orchestration and oversight roles here at home, where trust, institutional quality and proximity to decision-makers matter.
AI's equalising potential extends beyond white-collar work. A blue-collar worker who struggles with English could dictate in their mother tongue and have AI render as professional documentation in real time, freeing them to focus on their craft rather than their grammar. AI should be an equaliser that elevates the technical master, not a wedge that stratifies our workforce. AI tools can also power a new breed of local startups by enabling small hyper-efficient teams to create immense value and scale, achieving global reach with minimal manpower.
Singapore must therefore be at the forefront of this shift while ensuring that benefits accrue to all citizens. This will require workers and entrepreneurs who are trained, skilled and adept at harnessing AI tools and innovations and empowering their employees to do the same.
Our current efforts to reskill Singaporeans are often hampered by the trap of low-utility external training programmes which produce certifications that lack real-world currency in an AI-driven economy. These programmes enrich training providers while leaving workers with skills that have little economic value. This misalignment risks creating a two-speed economy where capital owners and tech-integrated firms leave behind those stuck in the slow lane of traditional employment, leading to a fundamental erosion of social cohesion and increasing the risk of long-term structural unemployment.
To address this, I propose the establishment of a National AI Equity Fund. This fund is a necessary safeguard to maintain the integrity of our social contract. It is a strategic surplus transfer from enterprises which benefit immensely from AI back to Singaporeans to facilitate our collective stability.
I will elaborate on the precise funding mechanisms shortly after I explain the uses of the fund. I propose the fund be organised in two distinct pillars.
The first is a social dividend where revenue is distributed as a direct payout to every adult Singapore citizen. I propose an initial citizen dividend of $500 per adult citizen, scaling upward as fund contributions grow. This is modest by design. It is not meant to replace income, but to provide a tangible signal that every Singaporean has ownership in our shared future.
Based on our current citizen population, this would cost approximately $1.5 billion annually – or less than 10% of last year's Budget surplus – and provide a meaningful return to every Singaporean household. This would serve as a social floor, ensuring that the gains from national digital prosperity provide tangible peace of mind and dignity for all.
This dividend will provide an additional cushion for families as the nature of work evolves. It also allows Singapore to reap the full productivity benefits of AI without overly exacerbating social inequality.
An argument could be made that the Community Development Council (CDC) vouchers already do this, but those are entirely discretionary. The social dividend I propose is a structural entitlement – a function of receipts rather than what the fiscal mood of the moment happens to be. That distinction matters enormously for a family planning its future.
The other portion of the fund will be dedicated to a mastery fund, which will be an employer-led, on-the-job training (OJT) model that moves training out of the classroom and into every enterprise.
I propose that the mastery fund provides a mastery apprentice wage covering 50% of the gross salary, capped at a median wage for six months, for any Singapore citizen entering or transitioning into an AI-augmented role. This rewards the worker's effort in adapting while lowering the barrier for firms to hire, train and retain talent in this volatile market.
Recognising that many SMEs lack the capacity to design structured OJT programmes, I propose that the fund also finance a pool of expert OJT consultants. These consultants experienced in OJT design will rotate between firms to structure OJT blueprints tailored to each firm's specific needs. This will help SMEs fill their talent gap while also addressing the need to create new steps in the ladder of training and apprenticeships for new entrants into the marketplace.
Furthermore, I suggest a mentorship credit be provided to employers to compensate senior staff for the time they spend on structured mentorship, turning our workplaces into true academies of mastery and ensuring that skills remain relevant to the actual needs of the economy.
The mastery fund should be made available to all business entities and societies that are founded and based in Singapore, including micro enterprises. The use of funds should be closely monitored to ensure that it genuinely contributes to AI mastery within each firm. I estimate the annual cost of the mastery fund to be approximately $1.42 billion.
Let me set out the financing details.
The first source is a marginal increase of two percentage points of the corporate income tax rate for firms with annual profits exceeding $100 million. By focusing on these companies, we capture the automation surplus from those best positioned to drive growth through AI rather than headcount. Whether global tech firms or traditional giants, these enterprises are at the forefront of decoupling revenue from labour. This tax increase will generate an estimated $1.5 billion annually, ensuring that gains from record-breaking efficiency are recycled back into the National AI Equity Fund for the benefit of all Singaporeans.
The second source is a targeted increase in the utilisation of our investment returns. I propose raising the maximum net investment returns taken into the Budget from 50% to 52.5%, with this additional 2.5% flowing directly into the fund. Based on current estimates, this would raise approximately $1.45 billion annually.
Our sovereign wealth entities, GIC and Temasek, have been early movers into the AI space, investing in foundational firms like Anthropic and committing billions to the AI infrastructure partnership alongside Microsoft, BlackRock and Nvidia. As these global Investments profit from the automation of labour worldwide, it is only right that we recycle a modest portion of those gains back into our own workforce.
Relocating 2.5% is not a radical request. It ensures our reserves provide more than just financial stability, but also the long-term economic agency of every Singaporean.
As we look toward the future, we cannot simply assume that displaced workers will transition smoothly into new roles as they have in previous technological revolutions. The steam engine did not replace human judgement, but AI may do just that. That is precisely why passive reskilling is insufficient and why the financial security of a social dividend is needed. Workers shifting towards less automatable roles in entrepreneurship, care work, the skilled trades, sports and the arts do not just need training, but time and security to make that leap.
Certainly, new jobs will emerge that we cannot yet imagine, but we must build a system robust enough to support our people even if that emergence is slower or more unevenly distributed than we would otherwise hope. The National AI Equity Fund provides a financial buffer for Singaporeans to make these transitions with confidence.
During this year's Committee of Supply Debate, I proposed a Youth Wage Credit Scheme – a targeted wage subsidy for employers who hire younger Singaporean workers. The National AI Equity Fund extends that logic into a broader longer-term framework for all Singaporeans navigating the AI transition and other technological disruptions.
Mr Speaker, the National AI Equity Fund is a renewal of our social contract for the digital age. We cannot allow AI to become a wedge that fractures our society. Instead, we must use it to become the greatest equaliser our nation has ever known. By establishing the social dividend and the mastery fund, we give every Singaporean a direct stake in our digital prosperity and the resources to stay ahead of the curve.
Let us make it our goal to ensure that as machines grow more capable, our people grow more secure. By acting now, we can ensure that technological progress serves the dignity and economic agency of every Singaporean. Sir, I support the Motion.
Mr Speaker : Ms Poh Li San.
1.43 pm Ms Poh Li San (Sembawang West) : Mr Speaker, in this Sitting and indeed, in the past few Sittings, few speeches have been made without mentioning the age of AI. Much has also been said by the Government about how the global disruption will impact our play, our work and our lives.
In policy-making, there is always the binary choice: ride this AI wave or be submerged and be left behind. This is Hobson’s choice and the answer is obvious. But we cannot ask a question of a strawman. Singapore is a small, open and digitally connected economy. AI will be a fact of economic life.
The Government has said that it will grow our economy, support our businesses and take care of our workers. But there is a difference between policy-making in the Ministries and implementation on the ground.
At street level, the AI transition looks intimidating, expensive, and for many middle-aged workers, a place of anxiety and confusion. This is the first wave of change and may be the hardest. The Government, businesses, unions and workers must struggle together to ride out this wave.
During the transition, some jobs will disappear, but new jobs will also emerge. Left to the market, there will only be growth where the strongest, fittest and most able benefit. In a city where the law of the economic jungle operates unfettered, let us be honest, the AI transition will benefit some, but not all. This growth will not lead to better and more prosperous lives for all.
But that is not the job for the market. That is our job; all of us here in this House must bend the market to our will: to create more high-value jobs for Singaporeans and retrain displaced workers but to do so in a way that also meets the business imperative for profits, so that our economy can continue to grow in the long run. We have built Singapore on solutions that met both interests in the past, and we must do so again in the future.
Let me talk concretely about two ways that AI can benefit our workers.
We used to say that people are Singapore’s only resource. We are now a superaged society with a total fertility rate of 0.87. Our resource pool is shrinking. Human resource is now the key bottleneck and cost driver for most businesses, especially SMEs which are hiring 70% of local workforce. If businesses close down, more workers will lose their jobs, even those not threatened by AI.
Our businesses have faced worsening labour shortages over the past decade. The 2025 ManpowerGroup's latest Talent Shortage Survey reveals that nearly four in five employers in the Asia Pacific region are struggling to find skilled talent, with 77% reporting difficulties. In particular, many jobs that Singaporeans cannot or will not do are done by non-Singaporeans.
But there is a limit to our dependence on foreign workers, including political and social constraints. We need AI-powered robots to replace foreign workers in roles that no Singaporeans would want to do or can do. For instance, heavy duty roles in construction, maritime and aviation sectors, that are exposed to the harsh elements of a warming climate. This will be a game changer for us.
The next frontier is in physical generative AI or embodied AI. In more recent times, generative AI technologies have been integrated with physical systems, enabling machines to interact with and adapt to the real world. It enables robots to learn complex tasks, such as manipulation and navigation, via simulation and transferring intelligence from digital models to real world hardware. Put simply, robots and humanoids capable of thinking and even perceiving, can be deployed in unstructured and dynamic environments, to assist or even replace human workers.
In recent months, companies such as Dexterity AI, Figure AI and Unitree Robotics have demonstrated their capabilities in AI-powered robots and humanoids in specialised roles.
Unlike generative AI tools like ChatGPT that trawl the Internet to train its models, physical generative AI tools need to be trained in contextualised environments for the roles and tasks they are set out to do. Over time, these physical AI capabilities will mature and become accessible to businesses facing manpower shortage. These AI-powered robots can help our businesses overcome manpower constraints, lower costs and increase profitability.
Physical AI robots are good at repetitive tasks but cannot replace every single role. Jobs and process redesigned into man-machine hybrid teams will be the new norm. Seniors and women can join the transformed workforce – with repetitive, heavy-duty roles done by robots and complex supervisory roles performed by humans. It will be a new model of freedom and empowerment, unimaginable today but a reality in the very near future.
Singaporeans can be upskilled as supervisors of robots. New high-value job roles such as design, build and maintenance of these AI robots will be created for young engineers and technicians.
More Singaporeans can retire at a later age if they wish to, since their roles will become less physically demanding. Seniors and women can join industries previously dominated by those with stronger physical abilities. And robots also do not carry any social baggage.
Mr Speaker, the transition to AI-powered robots is my area of work, and this is the vision we are working towards – to solve real problems for businesses and elevate the quality of life for workers. I feel strongly that our AI transition should be focused on customising physical generative AI solutions for our industries, so as to help every Singaporean on this journey of empowerment.
Prime Minister Lawrence Wong outlined four key pillars for our National AI Strategy. In particular, in the Advanced Manufacturing and Transport Connectivity sectors, AI-powered robots will indeed be the force-multiplier.
Are we ready for this transition? Not yet. We are well-positioned for it, but we must move fast. And I would suggest the following six steps.
One, unions should forecast which are the types of jobs and sectors at risk, as well as numbers of workers that may be displaced.
Two, MOM should fund retraining for affected workers, to prepare them for other roles or other industries.
Three, workers should also step up, learn new skills and be open to new job opportunities.
Four, the Ministry of Education (MOE) and IHLs should redesign academic programmes to shift students away from fields already taken over by AI. All IHL students should learn AI tools relevant for their disciplines.
Five, the Ministry of Trade and Industry should attract more world-class physical generative AI companies to set up headquarters in Singapore and attract talents for research and development (R&D).
Six, businesses should be open to work with AI companies to automate and redesign work processes, revamp job roles and create man and machine hybrid teams. Mr Speaker, I would like to share a few points in Malay.
( In Malay ) : [ Please refer to Vernacular Speech .] Mr Speaker, Singapore's AI transition must be managed jointly by Government, businesses, unions and workers to avoid uneven gains. Physical AI can ease labour shortages and reshape work into human machine teams. With job redesign, upskilling and education reform, workers can move into higher value roles, while businesses can grow with higher productivity.
To support this transition, Government support and regulation are crucial to fund retraining of workers, create higher value jobs, attract leading AI firms and ensure that AI is used ethically for broad societal benefit.
( In English ): In the near future, a new AI ecosystem will emerge. Technology companies create AI solutions, businesses own them and workers leverage them.
But the Government must set the rules. AI must be used as a force for good and not for criminal and harmful exploits. Establishing the ethics around AI use will make the difference between our society benefiting from the use of AI or becoming enslaved by it.
But there is also a deeper moral question relating to AI. AI is artificial; it has no intrinsic good, no value in and of itself. We, in this House, have a duty to bend the market in the use of AI, not just to forbid what is criminal but to enable what is fair, good and just. We must ensure that the AI transition does not merely create growth but creates jobs, benefits workers, strengthens businesses and elevates communities. Mr Speaker, I would also like to conclude my points in Mandarin.
( In Mandarin ) : [ Please refer to Vernacular Speech .] Mr Speaker, Singapore's AI shift needs a coordinated effort by Government, businesses, unions and workers to ensure that both businesses and employees stand to benefit, and to support and assist those who are affected. This includes providing retraining and expanding the scope of education, creating higher-value jobs so that more Singaporeans can adapt to these changes as quickly as possible.
Physical AI – robots combined with generative AI – can ease labour shortages by taking on tough and risky work in sectors, such as construction, aviation and maritime, while reshaping roles into human-machine teams.
I would like to propose several key steps to support the AI transformation:
First, companies should automate and redesign jobs; second, unions should flag at-risk roles; third, workers should keep learning; and fourth, Government should set clear AI rules. A national transformation should not merely be pursued for the sake of higher economic growth, but should ensure that all segments of society can benefit.
( In English ): Mr Speaker, the Motion on which I rise to speak today asks for us to affirm that AI transition "must not lead to jobless growth". And this "must" is not an empirical prediction, nor is it empty rhetoric. It is political resolve.
AI in this free market may or may not be the new model of freedom and empowerment for our people. It is our resolve that makes it so. Mr Speaker, I support the Motion. [ Applause. ]
Mr Speaker : Mr Andre Low.
1.57 pm Mr Low Wu Yang Andre (Non-Constituency Member) : Mr Speaker, the Motion before this House calls for an AI transition that does not leave Singapore's workers behind. The Prime Minister, the Labour Chief, the Government as a whole, have all said the same in the past months; that this is what they intend. What I want to examine this afternoon is whether the policy architecture we have is equal to the commitment we are being asked to affirm.
Mr Speaker, every AI deployment a firm makes is at its heart a choice. The firm can use AI to make its existing workers more capable, more productive, more valuable than they were before or it can use AI to do without those workers entirely. The economist's shorthand for this is augmentation as opposed to automation: augmentation where AI works alongside the worker, and automation where AI replaces them.
Stanford economist, Eric Brynjolfsson, one of the leading academic voices on AI and labour markets, has made a convincing case that in an unaided market without deliberate policies steering in the other direction, incentives systematically favour automation. Firms find it easier and cheaper to deploy AI to replace workers than to retrain them. The tax code, the labour market institutions, the cost structures of capital all tilt the playing field. Even though augmentation creates more total value over time, more good jobs, broader prosperity and a fairer distribution of the gains, the default trajectory of an unguided system is automation.
The Government's chosen and declared direction is augmentation. The Motion before us today assumes augmentation. The Labour Chief in this Chamber yesterday put the same commitment in his own words – not AI instead of workers but AI working for workers.
The philosophical direction is settled across the aisle. The substantive question is whether our policy architecture matches it.
There are three places where architecture is currently miscalibrated. Three places where, today, the system is permitting automation despite promises to the contrary.
The Labour Chief yesterday said that AI is also reshaping professional, manager and executive (PME) jobs in higher-end professions, like doctors, lawyers and accountants. The Prime Minister has said much the same – AI will affect Singapore's professionals, managers and technicians (PMETs) who have spent years building specialist careers and who are now being told that the ground below them is moving.
At his May Day rally last week, Prime Minister Lawrence Wong said, "We may not be able to protect every job, but we will protect every worker." The question is whether the instrument the Government has chosen, the SkillsFuture Job Seeker Support scheme, delivers on that promise. The Prime Minister has termed the Job Seeker Support scheme the "Singapore way", a more pragmatic, more Singaporean alternative to the redundancy insurance that is the Workers' Party's (WP's) preferred solution. That reads the Singapore tradition backwards.
Mr Speaker, the Labour Chief said in the Chamber yesterday that financial support during the transition is not welfare, it is an investment in worker outcomes. By that test, the tradition has long been built on investments of exactly that kind. The Central Provident Fund (CPF), MediShield Life, MediSave, these are all universal contributary schemes paid out when life's major contingencies hit. Each catches every worker because the contingency it insures against can hit every worker. That is the Singapore way.
The Jobseeker Support scheme is not built in that tradition. It is a tax-funded grant gated on pre-redundancy income, closer in design to a means-tested assistance than to insurance against contingency. As currently configured, it pays up to $6,000 over six months in tapering monthly instalments, starting at $1,500 and ending at $750 over the last three months and it is only available to workers who earn $5,000 a month or less before they were made redundant.
The Labour Chief acknowledged in this Chamber yesterday that the ceiling excludes PMEs who face the same displacement risk in the AI era and has proposed raising the qualifying ceiling to closer to the PME median gross income level.
If this proposal is adopted, it is movement in the direction that the WP has long argued for. But Mr Ng's proposal moves the line, ours would remove it. Raising the ceiling lets more workers into the scheme but it does not change what the scheme does for them. For those who do qualify under the ceiling, the taper carries its own message: a payment that starts high and slowly reduces is not a flaw. It is a countdown. And a countdown pushes a worker to take the first offer, not the right one.
MOM's own data tells us why this matters. Of retrenched residents in the final quarter of last year, 43.6% of PMETs had not found new employment within six months. That is the cohort that the Jobseeker Support scheme runs out on. And of those who do find work within six months, roughly four in 10 return at lower wages than before. So, they took what was available and not what their experience was worth. Most of us have experienced how the higher up the career ladder you climb, the longer it takes for you to find your next role.
Mr Speaker, a PMET who is pushed by a six-month countdown into a lower paid job that they did not want has experienced exactly the automation outcome that the Government's framework was supposed to prevent, with a small cushion attached for the fall. Raising the ceiling only widens the cohort, but it does not shorten the countdown.
Mr Speaker, the WP's proposal for redundancy insurance scheme is built in the actual Singapore tradition. We pay out 40% of last drawn salary with no income ceiling and no tapering mechanism. It is funded by employer-employee contributions in the same model as the CPF and it covers every worker who pays in, including the professionals the Labour Chief has identified as the most exposed because the contingency it insures against does not stop at $5,000, $7,600 or any other ceiling Parliament may set.
The Prime Minister said we must protect every worker. The instrument the Government has selected does not. The WP's does.
Mr Speaker, when a firm is contemplating a major AI deployment, it stands at a fork in the road: down one path it retains the existing workers and retrains them to operate alongside the AI; down the other, it retrenches, runs leaner, brings in a smaller AI-fluent workforce. The first is augmentation. The second is automation.
But what does our tax code say to the firm at that decision point? The current architecture rewards activity. It rewards capital expenditure on AI. It rewards expenditure on training. These are good things to reward, but what the architecture does not currently do is reward the choice itself.
A firm that retrenches its existing workers and trains a smaller set of new hires receive the same fiscal treatment as a firm that retain and retrains its existing workforce. A firm that buys AI to replace workers receives the same fiscal treatment as a firm that buys AI to augment them.
The tax code is silent at the fork.
And as Bronson previously observed, silence at the fork is not neutrality in consequence. When the tax code does not actively reward retention, the underlying economics tilts firms toward retrenchment. Labour is, after all, the most expensive line on the balance sheet and labour costs are permanent in a way that one-off training costs are not. An unaided market would choose retrenchment.
Yesterday, Mr Ng defended the CTC framework in this Chamber as the mechanism that ties enterprise transformation to worker progression and proposed expanding it through the new Tripartite Jobs Council.
CTC operates at the project level for firms that engage with it, with grant funding attached, expanding its reach scales the grant model, but does not change the broader fiscal architecture that every firm operates within, whether or not it is within the CTC scheme. And it is this broader fiscal architecture that shapes chief financial officers' financial decision-making at the decision point.
In February, in this House, I propose a retraining tax credit, a deduction available only to firms that can demonstrate that they have retained an existing worker into an AI-augmented role, rather than retrenching them. It is this missing conditional piece that will give firms a fiscal signal precisely at the point where they have to make a decision. This retraining tax credit would reward a proactive choice instead of simply investing in AI.
The fourth limb of this Motion affirms that economic progress must remain inclusive. That is a commitment about distribution, not just growth. My colleague Gerald Giam has proposed a National AI Equity Fund to deliver on that commitment structurally. The instrument I am proposing today is the diagnostic tool that any redistributive mechanism, including Mr Giam's, needs to operate on. Because the third condition for an augmentation strategy, to be real, is verification.
Mr Speaker, augmentation is, in the end, a testable claim. It makes a prediction that wages in the sectors where AI is being deployed alongside workers, will track the productivity gains that these workers helped to create. If that prediction holds, the framework that the Government has adopted is being delivered as advertised. If productivity rises in these sectors, but wages do not move alongside it, then what is being delivered is something other than augmentation, whatever language we use to describe it.
Right now, we have very few mechanisms and very few systematic ways of telling which is occurring.
The Government is investing serious public money at scale in four national AI mission sectors, advanced manufacturing connectivity, finance and healthcare. Public funds are flowing into these sectors and more through the CTC grants, the newly formed Tripartite Jobs Council, the Skills and Workforce Development Agency and various enterprise transformation programmes. These are appropriate investments, but public investment creates a corresponding public accountability obligation. where public money goes in, the public has a right to know what is coming out and to whom.
So, what I am asking for is a targeted transparency mechanism, an annual AI gains audit scoped specifically to the four national AI missions to start, reporting to Parliament on how productivity gains from state-backed AI investments are being distributed between wages and returns to capital. Over time, its scope and coverage can be expanded.
In February, in my Budget speech, I framed this as a distribution question. Today, with this Motion before the House asking us to affirm that economic progress must remain inclusive, I propose it again as something more fundamental. The AI gains audit is the most direct instrument available to Parliament to test whether the Government's chosen direction of augmentation is actually being delivered. If the gains are being shared with workers, the audit will say so and the framework will have evidence to back its claim. If they are not, we will know before the gap becomes a chasm and before this Motion becomes a statement of hope rather than of policy.
Mr Speaker, the choice between augmentation and automation is not made in one day. It is made every day by the architecture of the schemes we run, the tax code we maintain and the data we choose to collect. Whatever the House says today, that architecture will keep making the choice on our behalf.
Right now, my position is that the architecture pushes workers towards the first available job rather than the right one. Our tax code says nothing to affirm at the fork between retraining workers and retaining them and we have built no mechanism to tell whether the gains from public AI investment are reaching the people whose name that investment has been made. And that is why I support this Motion. I urge the Government to give it the architecture it requires so that we can make sure that no worker is left behind. Thank you, Mr Speaker.
Mr Speaker : Dr Hamid Razak.
2.10 pm Dr Hamid Razak (West Coast-Jurong West) : Mr Speaker, Sir, I declare that I am a business owner for a private orthopaedic practice, which is unionised, and also an advisor to the Health Services Employers Union (HSEU).
I rise in support of the Motion, an AI Transition with No Jobless Growth. AI is already here. It is not a pilot; it is already becoming a platform.
Sir, our question is not whether we adopt AI. We will. The broader question is whether we grow without leaving our workers behind. In the next decade, Singapore should be judged not by how fast we deploy AI, by how well we translate adoption into better jobs, better wages and stronger trust at the workplace.
I speak in three roles today as a professional, as a parent and as a Member of Parliament listening to my residents.
First, the professional anxiety. Many PMETs, they are not afraid of technology, they are uneasy about the uncertainty because AI rarely replaces the whole job. It unbundles tasks, it compresses teams, it changes what employers hire for. And when you cannot see how your role evolves, anxiety rises.
Second, the parental anxiety. Parents today ask very simple questions: will my child have a fair start? What will entry-level work look like? And if entry-level work shrinks, who will then train the next generation?
Third, the resident anxiety and this is the most practical one. Mid-career workers worry about displacement. Caregivers worry about time. Many people cannot just stop their work to train or retrain. They are not asking for guarantees. They are asking simply for a fair chance and a system that they can navigate.
Mr Speaker, Sir, we should be candid about AI. AI is smart, but it is not wise. It can hallucinate, it can sound confident and yet still be wrong. So, the future should not be about humans competing with AI. It should be about humans working with AI, with judgement, with verification and with accountability.
This Motion is just not about technology. It is about trust, job redesign and the worker journey in whole. Trust will determine if this adoption would succeed. If AI is experienced by workers as surveillance, the trust will thin and eventually break. And when trust breaks, adoption will slow and gains will not be sustainable.
Prime Minister Wong spoke about protecting every worker and scaling practical tripartite tools, like the CTCs for the AI transition. This is the direction we should double down on and I offer four practical moves on this front.
First, skills must be a pathway and not a menu. SkillsFuture is a major national asset. But on the ground, many workers tell me this. It is useful, but it is also overwhelming. Too many courses, too many badges, too little signal.
So, the problem now is not just access. It is navigation. A worker should not need to scroll for hours to guess what really matters to him or his next job. So, I suggest that we curate clearer AI-relevant pathways by sector, by job role, with a clear front door and clear employer recognition. And we can consider additional incentives for those who choose priority courses that support AI-enabled growth, especially when there is clear employer demand. This could mean higher subsidy tiers or outcome-linked support, such as completion plus interview, attachment or redeployment pathways subject to design and feasibility.
Second, tie AI adoption to job redesign. Many Members have spoken about this. If we fund adoption, we should ask, how will tasks change, how will workers be redeployed and how will the performance measurements continue to remain fair? Productivity must show up as better work and, ultimately, better wages for our workers and not only a shrinking headcount.
This is not meant to be punitive. It is meant to be practical. This is where our tripartite partners can help with playbooks, with templates and advisory support so that SMEs are not left alone to figure it out.
Third, bring AI readiness to the professional sectors, like clinics, law firms, accounting firms. Many are small, PMET-heavy and time poor. They want to adopt AI but they worry about safety, confidentiality, liability and trust.
One practical model already shared in this House is already emerging in healthcare. In April this year, HSEU and GP+ Co-operative signed an agreement to train primary care clinic staff in AI awareness, and to help primary care clinics adopt technology and redesign their workflows, supported by the CTC approach and the CTC Grants. I observed this partnership up close. The value is in making adoption practical, responsible and anchored on job redesign, and not just a tool roll-out.
I hope that we can extend this cluster-based, CTC-style approach to other professional sectors too, including law and accountancy, so smaller practices can move from uncertainty to readiness, with clear governance standards and worker protection.
For our workers, support should begin when need is recognised. That means faster job matching, modular training that fits real-life schedules and practical guidance for responsible transitions. In practice, structured coaching and a clear next step reduces anxiety, because it shifts a worker from waiting to acting. Mr Speaker, Sir, I will now speak in Tamil.
( In Tamil ) : [ Please refer to Vernacular Speech .] Hon Speaker, many feel concerned when talking about AI. They worry that jobs may disappear, that the value of skills may diminish and what may happen to the future of children. These concerns are real. The questions are also difficult, but our response is not to fear.
AI is advancing quickly, but human mercy, trust, sense of justice, creativity, language, culture – all these cannot be fully replaced by any machine. AI can compute; but the human connection, human judgement and the human responsibility will always remain with us. So, we must move ourselves from the path of fear to the path of opportunity. The rise of AI does not mean that humans are no longer needed. Rather, it makes it clearer on what truly are the important tasks that humans must do.
In this regard, the humanities are important; language skills are important; cultural nuances are important; social understanding is important. Healthcare, caregiving, education, social services, counselling and work that involves direct contact with people – these are fields that AI cannot replace.
The Tamil language and Tamil culture are a support in this endeavour; our literature nurtures the human feeling. Our culture strengthens social responsibility. That is our strength. So, this is what we must tell our young people: embrace AI but also grow the human capability.
Learning is not merely a certification. It is a path, a belief, a plan for the future. When growth comes, employment opportunities must come with it. Support for change must also come with it. So, do not fear. Let us have faith.
( In English ): Mr Speaker, Sir, no jobless growth must mean one thing. Growth that workers can feel, in wages, in dignity, in a clear next step.
So, I offer one governing standard. The test is not how many schemes we have. The test is whether a worker can quickly see the right course, for the right job, at the right time. Whether a parent can feel confident about their child's runway and future, and whether the citizen's journey feels seamless.
If we keep this direction and refine delivery with tripartite resolve, Singapore can deploy AI effectively, responsibly and at speed, while strengthening trust and protecting dignity. With these observations, Mr Speaker, I support the Motion.
Mr Speaker : Ms He Ting Ru.
2.19 pm Ms He Ting Ru (Sengkang) : Mr Speaker, Singapore's approach to AI is often cited by international institutions and consultancies like BCG, and prominent figures, such as International Monetary Fund (IMF) Managing Director, Kristalina Georgieva.
Our technological infrastructure and initiatives to upskill our workers are key parts of how we plan to confront the disruptions and opportunities presented to us by this new and rapidly developing technology.
However, we must also recognise and act on an additional uncomfortable reality. Singapore is one of the most vulnerable economies to AI disruption. International estimates suggest that around 60% of workers in advanced economies are in jobs that are highly exposed to AI. For Singapore, that share seems significantly higher. Because we are a high-skill services-oriented hub, estimates from the IMF indicate that approximately 77% of our local workforce is highly exposed to AI disruption, and our transition is likely to be sharper and more acute than in many other economies.
How is AI fundamentally reconfiguring our labour market? This can be understood through three distinct shifts.
First, many existing jobs will be transformed from within. AI is taking, or has already taken, over routine information processing tasks – drafting, summarising, extracting data and standardising analysis. Managers, health professionals and legal professionals are already using AI tools to handle these types of tasks, freeing up time for judgement, complex problem-solving and human interaction.
Second, some jobs will be displaced. In what economists call high exposure, low complementarity roles, AI can perform most of the core tasks on its own and there are fewer reasons to keep humans in the loop. Clerical support workers and many business and administrative associate professionals, whose work is built around routine documentation, basic processing and standardised customer queries, face the highest risk that their roles will shrink or even disappear. Advances in agentic AI technology and models have only sharpened this impact. In the UK, some financial institutions, like investment banks, have relooked their hiring of fresh graduates for certain roles because of AI's automation capabilities.
Third, AI will also create new jobs and new demands. We are already seeing rising demand for AI engineers, data scientists and AI-product specialists, but also for data-savvy professionals across finance, healthcare, logistics and education. These new roles tend to offer higher wages, but only for workers who can supply the right mix of technical and complementary human skills.
Yes, indeed, the AI job transformation is already here and we are in the midst of a major disruption. Yet, the impact will not be uniform across all professions, nor is it, and will it, affect our society and economy evenly. For now, AI disruption is strongest amongst white-collared workers, especially entry-level roles. Unlike previous technological disruptions that have historically affected blue-collared jobs, AI today will most affect cognitive, white-collared roles – a call centre agent, an admin officer or a junior business support executive, whose workday is built around standard processes, routine reports and scripted responses, is in a role where AI can perform almost all core tasks.
In such high exposure, low complementarity white-collar roles, employers can consolidate positions, slow hiring or redesign jobs to ensure that fewer people are expected to do more with AI as a simple justification. If we do not address this, the benefits of AI will end up with only a small group of workers.
Research suggests that productivity and wealth gains could disproportionately accrue to those best positioned to leverage AI capabilities. One documented economic effect of high-skill job creation is increased local service demand. Studies from major tech hubs, including San Francisco, indicate that each high-end job is associated with the creation of approximately four jobs in local service sectors, such as retail and food services.
Even if such spillover effects generate more jobs, the quality and availability of these jobs for vulnerable workers is less certain to me. In Singapore, lower wage and routine-intensive roles are more likely to be held by vulnerable worker groups who may also face greater displacement risk from automation.
International institutions, including the IMF and World Bank, have noted that AI could exacerbate income inequality in the absence of policy intervention. The extent to which spillover effects from AI-driven growth would benefit lower-income workers remains uncertain. We need Singapore-specific research, modelling these distributional impacts and to make this data publicly available, to inform more targeted policy responses.
We must also remember that Singaporeans are already feeling the strain of rising property prices and higher cost for essential services. These pressures are real. They have been building for some time as we are a small, open economy, significantly dependent on capital inflows. Would AI's effects drive further unequal wealth accumulation? It is, therefore, a fair and pressing question to ask: could AI-driven economic activity inadvertently add to daily cost pressures?
Beyond broad economic pressures, we must turn toward the human face of this transition. As jobs continue to be reshaped and workers continue to be upskilled, we cannot leave behind those who face systemic barriers as our nation progresses towards an AI-ready future. Amongst them are persons with disabilities, women, lower-income Singaporeans, as well as young graduates.
AI can introduce new forms of discrimination against persons with disabilities. As AI algorithms are often trained via pattern recognition, they arrive at determinations based on common patterns within datasets. Thus, if skilled historical data is being used to train AI for, for example, recruitment processes, AI might reinforce this bias for job applications from persons with disabilities and any other group which historically, is not well represented in this space.
Female workers, too, face a heightened risk of marginalisation from AI. A 2024 IMF report on Singapore's labour market found that women are under-represented in AI-intensive science, technology, engineering and mathematics (STEM) roles, and among workers with AI engineering skills. Women in STEM held 29% of entry-level positions, 24.4% of managerial positions but only 12.12% of C-suite roles. Altogether, this means that they are less represented in what is regarded as the safe side of AI. They would thus be less well positioned to benefit where AI compliments high-skill work. Additionally, International Labour Organization data, released in March 2026, found that occupations dominated by women are nearly twice as likely to be exposed to GenAI risks compared with male-dominated ones, with even stronger differences presenting when looking at high automation risk.
Taken together, this creates a double disadvantage. Female workers are less likely to gain from AI's benefits, while remaining more vulnerable to displacement. In short, women face higher risk and have fewer opportunities.
For our young graduates, new uncertainty has been introduced by the way AI has been reshaping jobs. The erosion of entry-level jobs has presented a catch-22 dilemma for Generation Zs. While companies are still looking to hire professionals for experienced roles, young graduates have fewer opportunities to gain such experiences as jobs are absorbed by AI.
With more than 20% of graduates unable to secure full-time permanent roles in 2025, a nearly 5% increase from 2023, and over 60% of graduates claiming that the job search has become difficult, it is only natural that young graduates have become more anxious about landing full-time employment.
Adding to this is recent research, which has shown that simply being aware of AI's potential to augment or threaten one's job can increase burn-out, mainly by heightening job insecurity and emotional exhaustion amongst workers.
While AI is often associated with disruption to white-collar work, vulnerable workers and families face significant risks too.
Unequal access to AI tools and training could entrench existing disadvantages. Those without the resources of home or home environments conducive to learning new skills may find themselves falling further behind. If left unaddressed, then this risks hardening inequality across generations.
Measuring impact on Singapore beyond economic output, what does this all mean for Singapore? Just this past week, we have seen the launch of the Marriage and Parenthood Reset Work Group. What is the effect of AI advances on our birth rate? Economic insecurity has already been cited by young Singaporeans as a reason for delaying or foregoing parenthood, but the barriers go beyond finances. Job uncertainty erodes a sense of stability and confidence in the future, the feeling that one has a firm enough footing to build a family and put down roots. If AI-driven disruption deepens this broader sense of insecurity, we can reasonably expect further downward pressure on our already tragically low total fertility rate.
The Government has to be even more targeted in ensuring that all workers, regardless of their gender, age, occupation, income and accessibility needs, are fully prepared for the disruption caused by AI, to ease financial pressures on vulnerable workers who are made redundant. This will minimise the uncertainty and toll of unemployment on both workers and their families as AI displacement becomes more commonplace.
To ensure our policies are working, we need more public data for us to measure AI-driven disruptions on our labour market. For example, how we measure the success of our AI programmes.
Following up in response to my Parliamentary Question on 24 February this year, I noted that the AI apprenticeship programme is currently assessed through three primary indicators: one, the total number of practitioners trained; two, the percentage of AI Apprenticeship Programme graduates who took up AI engineering related roles, and; three, the completion and supervision of project quality. It is a good start, but they do not tell us the effects of AI programmes and disruptions on different groups of society. These measures focus on throughput rather than equity. We need data on wage trajectories, job quality and retention in AI roles two to three years after the programme is completed. We need more data.
First, the participant data profile should have more details made public. This can include previous occupations, income band before training, age, gender, education and disability status. This allows us to see where participants come from. High exposure, low complementarity roles, or already high complementarity roles. This will inform if vulnerable groups are even putting their feet through the door.
Next, more accurately measuring AI disruption in the wider labour market can come in the form of exposure complementarity mapping, thus understanding whether jobs are high exposure and low complementarity, and to establish a tuned framework to track displacement, wage changes and job quality across demographic groups. Such data gives the Government a clearer picture of how AI is affecting different communities, so that support can be directed where it is most needed.
I will now turn to some thoughts on how our youths can address the challenges of AI.
If AI displaces a significant share of entry level roles, young workers may find fewer opportunities to build the foundational experience traditionally needed to progress into senior positions. One of our nation's solutions could be to better encourage and support entrepreneurship amongst youths. This will allow them to also gain valuable skills independently, rather than wait to be picked up to be employed by an established firm.
This approach builds on an already open door. AI has greatly reduced barriers to starting a business by being deployed to build websites, analyse data, run marketing and even automate back-office tasks. We have many schemes for startups, such as grants and boot camps, but do these initiatives adequately provide sustained long-term support across the full lifecycle of a burgeoning firm?
Moreover, our grant architecture remains milestone heavy and programme bound, encouraging compliance over competition. We need a culture and framework that recognise the value of a failed startup, or that support founder-led networks over time. Drawing on lessons from other entrepreneurial hubs, what are the areas that have inhibited Singapore's ability to establish a more sustainable ecosystem conducive for entrepreneurs?
First, we must continue to build sustained informal networks that made an entrepreneurship culture self-sustaining. Our current networks are often programme based and time limited, skewed towards short-term coaching. Yet research shows that informal mentorships arising from mutual choice and affinity are far more effective than administrative matching. If mentorship is only linked to short-term grants, our youths may miss out on the benefits that accrue from the trust-based guidance that can be seen in, for example, Silicon Valley. In leading entrepreneurial hubs like Silicon Valley and Shenzhen, informal founder networks have been a critical but often overlooked driver of success. They enable knowledge sharing, supply chain connections and a spin-off of new ventures from anchor firms.
Singapore can gain much from this. While we have anchors like Grab and Block 71, the Asian Development Bank has noted that our ecosystem trails others because our collaboration remains policy-driven rather than organically clustered. How can we reduce administrative burden on founders to ensure that they do not become overly occupied with meeting grant milestones instead of establishing the market competitiveness they need to survive AI-driven disruption.
One possibility is to limit formal reporting to end of grant rather than more regularly to strike a balance. Singapore must also better leverage our anchor firms. Companies like Grab, Sea and Singtel hold deep reservoirs of technical expertise and industry networks that largely remain locked within the firm.
Could we use targeted tax credits or co-investment matching for peer development programmes to encourage anchor firms to run structured mentorship and spin-out programmes for early-stage founders? This will allow organic networks to form around existing reservoirs of excellence, rather than hope that Government grant cycles will do so.
The private sector must lead, and the Government's role should shift from convener and gatekeeper to catalyst. This is how we can start to grow our entrepreneurial system from within industry.
We must also learn to value failure. Singapore's culture of economic emphasis and social conformity makes us sometimes afraid to fail. A 2018 piece of study by the Organisation of Economic Cooperation and Development (OECD) found that Singapore students expressed a greater fear of failure than their peers compared with any other participating country. Yet entrepreneurship means being tolerant of failure. Founders have to make decisions with incomplete information and meaningful innovation has to be backed up by some freedom to fail. We have to treat failure as a stepping stone rather than a stigma, or we end up stifling the very ecosystem we are trying to build and leave our youths ill-equipped to flourish in an age of disruption.
We can do so by beginning our own transition towards a better space for entrepreneurship, encourage experimentation and normalise entrepreneurial failure as growth and experience.
Failure should be a stepping stone, not a dead end. And we began this transition within schools where we have to move away from perfect scores, where we have entrepreneurial projects in schools which expose students to the inner workings of a startup. We should also showcase failed projects for their bonus.
Singapore's current bankruptcy framework can also be re-examined to better support entrepreneurs. Currently, founders who fail face the same restrictions as any other bankrupt, with travel bans, director disqualification and no automatic discharge, regardless of whether their failure was the result of genuine risk-taking or financial misconduct. Could we explore a dedicated pathway for bona fide startup failure, one that allows founders to be discharged sooner, resume directorship more quickly and have their experience recognised as something valuable rather than a liability. It is not to make failure consequence free, but to ensure that the cost of an honest bet gone wrong does not permanently deter our most enterprising young Singaporeans from trying again.
Finally, it is also my hope that we use our experiences navigating the AI transition to play a regional and global role as other economies too attempt to navigate the disruption.
Singapore comes from a place of strength and we are already intentionally deciding to lead the way when it comes to setting the agenda in global AI governance. Our stewardship role must extend beyond frameworks, and we have to play our part in addressing global imbalances in AI development and use reflected in recent data. World Bank 2025 data show that high income countries account for 87% of notable AI models, 86% of AI startups and 91% of venture fund capital funding, despite representing just 17% of the global population. There is justifiable concern about how vulnerable groups and the global South are woefully underrepresented in the AI space. As responsible world citizens, we can do our part to address this.
Recently, we have already begun developing AI tools tailored to Southeast Asian languages through Project Sea Lion, recognising that much of the developing world risks being left behind by AI systems built on Western data. We should build on this by championing equitable AI access across the Association of Southeast Asian Nations (ASEAN), exporting our governance expertise to nations that lack the capacity to develop their own frameworks and ensuring that the rules governing AI reflect not just the interests of the powerful, but the needs of the many.
This is not merely an abstract foreign policy and ambition. It has direct consequences for jobs here at home. Singapore's standing in the global AI ecosystem, gives us leverage to shape how AI tools are built, deployed and adopted across the region. We should use that leverage intentionally. When our researchers —
Mr Speaker : Ms He, you have a minute left.
Ms He Ting Ru : When our researchers develop AI system that work across Southeast Asian languages, we create tools that can be deployed in our own service sectors, our hospitals, our schools. When our companies lead in AI adoption, we generate demand for new skills, new roles and new industries that our workers can be trained into.
We must ensure Singaporeans are in the room where these technologies are being built and not merely be on the receiving end of decisions made elsewhere. Our global AI leadership is ultimately an investment in ensuring that the answer is the former. This approach also ultimately has the added benefit of creating more jobs and opportunities for Singapore in what would be a true trickle-down effect. I support the Motion.
Mr Speaker : Mr Yip Hon Weng.
2.39 pm Mr Yip Hon Weng (Yio Chu Kang) : Mr Speaker, Sir, I declare I work in a global investment firm working on ecosystem workforce strategies.
In recent weeks, a plane has been circling San Francisco with a banner that reads, "Stop Hiring Humans." The same message appears on billboards and bus shelters across the city, alongside slogans such as "The Era of AI Employees Is Here." The campaign is the work of Artisan, an AI startup. This is not just a marketing stunt. It reflects a fear that the future of work may exclude people rather than empower them.
I rise in support of this Motion because that must not be Singapore’s approach. AI must not be a signal to workers that they are dispensable. In my work in Temasek, I have seen how technology disrupts industries, and I want to lay out my key thesis upfront. To achieve growth without casualties, enterprise AI adoption cannot just be about buying technology. It must follow a deliberate thread: we must first build AI fluency; use that fluency to drive workflow and job redesign; and ensure that this redesign leads to tangible, shared outcomes for our workers.
Let me begin with an important observation. In the Temasek ecosystem, many companies are already investing in AI. Tools are being deployed and pilots are multiplying. But the real constraint we are seeing is not technology, compute, or capital. It is workforce readiness. We are not short of technology. We are short of transformation.
In our AI Fluency workshops with over 20 Temasek portfolio companies, we see clearly that fragmented AI literacy remains a primary bottleneck. We are working closely with Chief Human Resource Officers and Chief Technology Officers to bridge the gap between adoption and actual value creation. Fundamentally, this is a skills-matching problem. Where skill supply lags, opportunity does not disappear. It simply moves elsewhere.
The AI transition we are debating is structural, global and accelerating. Tasks that took weeks now take hours, and soon minutes or seconds. Change no longer moves in a linear way, it moves exponentially. In this context, adoption is no longer optional. In Singapore, while large firms wrestle with legacy systems and heavy compliance, and SMEs face severe constraints in capital and bandwidth, the implication is the same for both: enterprises that do not adopt AI will struggle to remain competitive.
But if AI adoption is necessary, disruption is inevitable. We must be clear about the risks to workers if this transition is not managed carefully. On one hand, as AI lowers the cost of many tasks, demand for those tasks can expand rather than contract. Economists refer to this as the "Jevons employment effect", where efficiency leads not to less work, but to more work in new forms. We have seen this before. ATMs reduced routine tasks but expanded banking. Word processors increased output and shifted work to higher-value roles. AI will likely follow the same pattern.
But the practical reality often produces a K-shaped outcome. Experienced, AI-enabled workers capture disproportionate gains, while those without such capabilities, especially entry-level workers, risk falling behind. So, the question is not whether AI creates growth. It is who that growth accrues to. The real risk is not that AI replaces jobs. It is that it replaces opportunity at scale. A worker may remain employed but face slower progression and a quiet erosion of experience. Our task is not to deny disruption. Our task is to govern it.
Mr Speaker, Sir, if we are to govern this disruption effectively, the debate must shift. It is not enough to observe whether firms adopt AI. We must demand accountability for what happens after adoption. The question is simple. Are workers better off after transformation than before it? We must ask: are jobs and workflows being redesigned? Are gains being shared? If AI raises output but weakens livelihoods, that is not transformation. It is exclusion. We must ensure public funds do not subsidise this.
Hence, I ask the Government: can we establish clear conditionalities for our support schemes? If public grants are funding a company’s pivot, should it not be explicitly tied to a national, human-centric scorecard? A scorecard that tracks the number of net new roles created, the scale of workflow redesign, wage improvements, staff retention and upskilling. If we are serious about no jobless growth, our adoption metrics must move beyond counting jobs to measuring careers.
From our work in AI fluency in Temasek, we know companies are urgently asking for guidance in leadership capability, job redesign, measurement of outcomes, trust and governance. Companies cannot navigate this alone.
This is where our Labour Movement, NTUC, e2i and the unions come into play. We must empower them to provide this guidance, ensuring that union leaders and management sit at the same table, mapping out the enterprises' technology roadmap and the workers' retraining schedules simultaneously. If companies need guidance on job redesign and shared outcomes, our tripartite partners must be right there on the ground with them.
The recently announced Tripartite Jobs Council is a timely step, but it must actively bridge the gap between firm-level transformation and worker-level outcomes. Because the real challenge is not introducing AI into firms, it is integrating workers into that transformation.
This brings us to the central challenge of this transition: AI is not just a technology multiplier; it is a leadership multiplier.
Leaders must have real conviction in the transformational potential of AI and lead by example in using it, guiding its application and bringing their workforce along. AI can write, design and optimise. But it cannot exercise judgement, build trust or carry people through uncertainty. That responsibility remains human.
Mr Speaker, Sir, employers have a critical responsibility, but without the right leadership, the easiest way to adopt AI is to simply cut headcount. We are already seeing this tension play out globally. We have seen Amazon announcing 16,000 job cuts in early 2026 while leaning further into AI for corporate efficiency.
When AI is introduced mainly as a headcount strategy, it breeds fear. When introduced as a capability strategy, it builds trust. Stories of this fear have long played out in our workplace, even among AI-enabled workers.
When generative AI first started trending, some workers proactively explored AI on their own to improve their efficiency at work, but kept it a secret. They worry that if they reveal the source of their newfound productivity, they will eventually be made redundant or be loaded with more work without extra compensation. When trust is absent and gains are not shared, workers hide their capabilities rather than sharing them.
True transformation requires leaders to adopt a long-term perspective. Firms must expect short-term productivity lags as workers transition to new ways of working and they must create space for experimentation. However, we must acknowledge that many businesses struggling with high operating costs in this challenging economy do not have the luxury of time.
I ask the Ministry: how can we better support companies to absorb these short-term lags, ensuring that the cost of time required to retrain workers does not become an excuse for lay-offs?
Mr Speaker, Sir, as employers lead this transition, there is one structural shift we must recognise clearly – before job redesign can happen, we must have workflow redesign.
We ask workers to change, but we leave the system unchanged. AI cuts across functions and domains, reimagining how processes connect and how value is created. That reshaping of the entire process must come before individual roles themselves are redesigned.
However, here lies a significant gap in our current policy approach. Today, much of our national support is heavily focused on individual job redesign. We ask workers to adapt to new job scopes, but legacy company processes are left unchanged.
As a result, AI is often layered onto outdated workflows with silos and fragmented data. Productivity stalls and frustration rises. Workers resist change not because they are stubborn but because it is deeply frustrating to use advanced AI tools within broken workflows.
So, I ask the Government and our tripartite partners: can we expand our support schemes to explicitly look into workflow redesign? How can we provide enterprises with the expertise and funding to reimagine their cross-functional processes first? If we fix the workflow, workers will naturally see the value of the technology, turning inertia into eagerness to adapt.
Mr Speaker, Sir, even with the best workflows, disruption will occur. Some workers will be displaced and some roles will change faster than expected. We often point to existing measures like SkillsFuture and career conversion programmes. But here is a hard truth – if these measures are sufficient, why are our workers still so deeply anxious?
The answer is that AI disruption moves at an unprecedented speed and workers worry our safety nets cannot catch them fast enough.
I ask the Government: how are we rigorously tracking the speed and effectiveness of our existing measures, particularly the financial runway for displaced workers, to guarantee that they remain truly responsive?
For the workers who remain, basic AI capability will no longer be a distinct advantage. It will be the price of staying in the game. The task is to help workers transition from being mere AI users to becoming AI conductors, workers who know how to curate, steer and verify AI outputs.
This brings me to a critical concern regarding our workforce pipeline. If AI automates drafting, summarising and first-pass analysis, what happens to our entry-level jobs? If young graduates cannot get a real first job and the mentorship they need, they will never gain the foundational experience earlier cohorts relied on to grow and we risk losing our future workforce.
I ask the Government: how are we working with employers and industry leaders to protect and redesign entry-level pathways so that our youth can develop the professional judgment required to become the AI conductors of tomorrow?
In conclusion, Mr Speaker, Sir, let me return to where I began. In the Temasek ecosystem, we see companies investing in AI, piloting new tools and pushing forward with transformation. But the decisive constraint is not the technology. It is whether the workforce is ready.
I shared this story with chief technology officers across our Temasek companies. A factory invested heavily in new machines. They were faster, smarter and more efficient. But productivity fell, not because the technology failed, but because the people were left behind.
On another line, the company did something different. Instead of replacing workers, they retrained them. Machine operators became data readers. Technicians became problem solvers. What changed was not the equipment. It was the capability and confidence of the people using it. Soon, breakdowns fell. Ideas came from the shop floor. Workers who once feared change began to lead it. The same machines. The same factory. But a very different future.
That is the lesson. Technology may set the pace. But people determine the direction. And in an AI transition, fluency, not just adoption, determines whether that direction is inclusive.
That brings us to the Motion before us. An AI transition with no jobless growth is not just about creating jobs. It is about ensuring that workers advance with technology, not fall behind it. It is about translating productivity into progression. It is about turning innovation into shared outcomes.
If adoption builds capability, then fluency must build advantage. If work is redesigned, then skills must be deepened. If growth is created, then it must be broadly shared.
But this will not happen on its own. It requires coordination. It requires leadership. It requires trust. And this is where Singapore has a unique advantage. Our model of tripartism, where Government, employers and workers move together, gives us the ability not just to react to change, but to shape it. This is our secret weapon.
When firms invest, workers must be equipped. When jobs are redesigned, workers must be involved. When disruption occurs, support must be credible. Because this is not just a technology transition, it is a workforce transition. The establishment of the Tripartite Jobs Council is an important step in ensuring that this alignment happens in practice.
Technology will move. Markets will adapt. But we must be clear about the future we are building. It cannot be one that says, "Stop having humans". It must be one that says, "Invest in people". Whether our workers advance is a choice we must make together, deliberately and decisively. Thank you, and I support the Motion. [ Applause. ]
Mr Speaker : Assoc Prof Jamus Lim.
2.54 pm Assoc Prof Jamus Jerome Lim (Sengkang) : Mr Speaker, my contribution in today's debate is straightforward. I argue that if our objective is to protect our workforce from job losses that could result from an economy-wide embrace of AI, then our efforts today should mainly be directed towards policies that promote new hiring rather than those that focus either on reducing displacement or pushing for retraining.
The explanation is simple. The evidence shows that to date, job displacement of current workers due to AI has been modest and localised whereas the hiring slowdown of new workers is already evident and even likely to accelerate.
In principle, the consequences of AI could bite on two ends of the labour market. It could reduce the incentives for firms to hire, thereby lowering the number of total new job openings, or it could raise the frequency of hiring by companies or induce workers to quit.
Hiring slowdown results when AI tools allow a company to cheaply and effectively replace job functions that they previously needed to hire a human worker for. This is especially the case for entry level since the lack of experience among new graduates and the relatively straightforward grunt work new hires are generally tasked to do make this group somewhat less valuable and more replaceable by a machine.
But as many observers, including Members of this House, have already pointed out, this is a chicken and egg problem. If we do not absorb new workers into our corporations, we surely cannot expect them to gain the necessary experience and job-specific skills that make them valuable as mid-career professionals.
Job displacement occurs, in contrast, when AI tools reveal that certain roles are no longer needed as they can be well replicated by AI. Tasks that used to be done by a human are replaced and if there is nowhere else in the firm for this person to be reassigned to or if the individual turns out to be too costly, then they are let go.
On the positive side, AI could open up new opportunities for displaced workers to pursue a career elsewhere, either because they gain AI-related skills that make them more valuable in the marketplace or because perhaps they could go into business for themselves.
For now, however, the message from studies worldwide is clear. While there has been little evidence of displacement thus far, there are more ample signs that there has been a decline in hiring. This has been the case for AI-exposed sectors following the advent of generative AI such as ChatGPT and is likely to accelerate as agentic AI matures. Prospects are especially perilous for early-career, entry-level workers. And even when hired, such workers tend to receive lower salaries.
The reasons for this are intuitive. AI mainly substitutes for mechanical, repeatable and well-defined tasks, which are mostly performed by junior employees. Firms still value the maturity and experience of senior employees and, by and large, would rather skimp on hiring and re-allocate their otherwise loyal staff rather than giving them the boot.
Sir, these trends are also visible in our local labour markets. Thus far, AI is yet to contribute much to job displacement here. MOM's latest labour market report reveals that since 2023, overall retrenchments have remained stable and the unemployment rate of about 2% has not budged much since 2022.
In response to questions about the PMET sector, Senior Parliamentary Secretary for Manpower Shawn Huang also pointed out that retrenchments among PMETs in AI-exposed sectors such as finance and infocomm remain low, numbering only 960 in the final quarter of last year. Moreover, Senior Parliamentary Secretary Huang also pointed to the large number of vacancies – around 10 times higher – in these sectors over the same period. This could be interpreted as a healthy robust labour market for new hires, but I will caution against it. This is because as any jobseeker will tell you – an opening does not a job make. These jobs have to be filled, ideally by Singaporeans who are themselves looking for jobs.
Here is where the picture is less encouraging. The latest graduate employment survey shows a drop in the share of graduates that managed to land a job in almost every single field of study, with around one in every four graduates unable to secure full-time employment.
In response to a Parliamentary Question filed, Minister Desmond Lee pointed to how the decline was due to a post-pandemic hiring surge and that what we are seeing is simply a mean reversion to the early trend.
I am less sanguine. Based on my calculations, what is true that pre-pandemic and permanent employment among graduates averaged around 70% in 2020, it was closer to 85% as recently as a decade ago, which is significantly better.
The youth unemployment rate which is lower in the 2020s then it was in the prior two decades has admittedly, elsewhere in the world, also steadily inched up since 2022, by around a percentage point. And unlike earlier instances when uptakes followed an economic cycle, Singapore's economy is actually currently in expansion.
On the demand side there has also been reports that some employers have been tentative and reluctant to onboard workers, albeit these thankfully appear to be a minority for now. Moreover, this pessimistic picture also mass more troubling pathologies. Many Singaporean workers may have had to contend themselves with opportunities that do not fully employ their skills and talent. Such mismatches are not well captured by the aggregate data.
Think of the graduate with an advanced degree from a local university but nevertheless still felt compelled to work in food delivery, or the student who spent years at a top university abroad but has been repeatedly rejected by employers since coming home. Or the experienced mid-career professional with bills and a young family to support, who floundered unsuccessfully in the job market month after month despite upskilling as advised. I believe others in this House will have seen cases similar to these in our weekly Meet-the-People Sessions. And indeed, two recent studies published by NTUC and MOM, corroborate these examples.
Overqualification is the greatest for those who are early in their careers and the gap between the involuntary and voluntarily underemployed is greatest for those under 30. Coupled with the global climate of economic policy uncertainty, we may be heading toward an era of the so-called "great hesitation" in hiring in our local labour markets, similar to what has been observed elsewhere in the world.
If our goal is, as the title of the Motion suggests, to avoid jobless growth, then it follows that we should prioritise policies that target the hiring of the labour market. Let me offer a few.
First, we can improve the incentives for companies to hire fresh graduates. As I have shared in the long cut for this year's Committee of Supply, this would call for expanding the existing GRaduate Industry Traineeships (GRIT) programme to a national level, cross-sectoral national internship initiative. Young workers will be free to apply their SkillsFuture credits toward paid apprenticeship and internship programmes with companies willing to take them on.
Corporations, especially SMEs, should also be able to submit credible proposals for in-house, on-the-job training to MOM, which will then offset the cost of taking on these trainees, drawing on the SkillsFuture enterprise credit and other subsidy schemes already earmarked for businesses. My hon friend, Gerald Giam, has also proposed an AI mastery fund for this purpose, which is complementary to what I am suggesting here.
Second, such short term, by which I mean six months to a year, apprenticeships and internships should also embed a clear employment pathway, conditional on reasonable performance on the part of the employee, unless a waiver is granted to employers due to changed economic circumstances.
These trainees should be treated as employees under the Employment Act and receive the same legal protections and entitlements, including a minimum period of annual leave, which GRIT trainees currently do not receive.
Third, we can ramp up the delivery of social skills training in communication, empathy, judgement, networking and vision in the final year of their tertiary education prior to workforce entry. Research has shown that AI is the most complementary to workers when the job demands require the fulfilment of not only cognitive tasks, but also in iterative collaboration between humans and AI. But our graduates often load their school time with the pursuit of academic competencies, leaving them woefully under prepared for such interfacing functions.
Fourth, if indeed we stand by our belief that we want our graduates to focus on acquiring competencies rather than certifications, as MOE has made clear in its support for stackable, micro-credentials pathways in our autonomous universities and as corroborated by recent research, then we should put our money where our mouth is and end hiring requirements that insist on a diploma or degree in the public sector if the competency can be demonstrated, otherwise. This can occur with proof of skills via a series of micro credentials, or when candidates pass a live demonstration during the interview stage.
Sir, AI is a general-purpose technology. Like all general-purpose technologies before it, AI will destroy perhaps as many jobs as it creates, but as we confront the bleeding edge of this transition, we must set the stage for those who are most affected by the roll-out, which, for now, at least are clearly our young entry level workers. We need targeted policies that will help sidestep the great hesitation in hiring them. This is how we best ensure that the growth promises of AI are not overshadowed by fears of millions of missing jobs.
Mr Speaker : Dr Neo Kok Beng.
3.06 pm Dr Neo Kok Beng (Nominated Member) : Mr Speaker, Sir, I used to be a visiting professor of innovation policy at Harvard Kennedy School for 10 years and currently I am still a visiting professor of innovation management at Fudan University.
AI is what we really call disruptive technologies, of which I spent my doctorate on. It is paradigm shift and when you say paradigm shift means it is shifting from one, or completely one industry to other areas that actually will destroy the old ways of working.
Let me just give you an example. Last week, I visited at AI startup in the Science Park, so we were talking about collaborations on a couple of projects in media, and I only saw two persons in the room with a capacity of 20 persons. So, I asked the chief executive officer, "Where are all the staff?" They say, well, they never come to the office, but they are working. And I say, "When do you see them?" They say, "Well, I saw them last week when the network went down and there were no communications. As the broadband was down, they came back to check on their agents to see are they really working?" So, I asked, "How would you know they are really working?" The answer, "Well, actually, they produce the code at any time, all the way, 24/7." So, he has got agents, they are working 24/7. So, I said, "How do you measure the productivity?" And I asked one question, "How many tasks or jobs have you replaced since, or how many jobs have you done, the capabilities of people doing the engineers or the coding jobs, since last year, one year ago?" And the answer that came back is five. So, compared with last year and this year, this guy has agents working for him and he performed the task of five persons.
We can look at it from two angles: one is this company is supremely productive; the other thing to look at is, we have five jobs less.
Which one would you choose? Well, if you are the CEO, you know what to choose. If you are the NTUC Secretary-General, I am not so sure.
So, I asked the next questions. He happened to be interviewing potential staff for his expansion, and so I ask him, "How do you find this guy who is doing his Masters in computing? Is he up to par?" And the answer that came back is, "Well, based on what I discussed with him and his answer shows he is two years behind the current technology".
Two years behind and I was like, wow, so when this guy graduates, maybe next year, what is going to happen to him? Technology is always two years in front of him.
And therefore, that means, do we put this guy into the real environment, which is basically internship, on-the-job attachments, or this person when he graduates, he really needs to speed up on the current competencies where the AI technologies are moving so fast. So, really, this is a tsunami.
Personally, I get myself involved in the small little non-governmental organisations. We are working on one pet project, which is to monitor the senior citizens who are staying alone, so that if anything happens to them, we know. And we actually are using AI, because most of these senior citizens speak dialect dialects, so we are using robot companions, small ones, with the ability to understand dialects; and to monitor them. It is very good, because it is very difficult to get caregivers to go around monitoring and visiting these people.
So that is one thing that the AI is really, really useful for. The robotics AIs are really useful so that we can cover things that Singaporeans do not want or are probably not so suitable to do the job.
The other projects that I am working on is imaging, using AI for magnetic resonance imaging. We will talk about it in some other time.
The issue is that the workplace experience is now changing quite a bit. So, do we, or our staff, or our existing people, or PMEs, or workers have the skills for them to carry on in the job. I am glad actually that the Labour Movement has introduced the CTCs. It is a very good mechanism to bring AI into the workplace or working with the enterprise. And I am glad that there are such grants for it.
But the question then is how do we define competencies? Some Members talked about the workflow redesign, process redesign – but how do we know that at the end they are competent? Where are the competencies level at the workplace? So, I think the new agency, SWDA, should be able to work together with the Labour Movement to define the competencies.
But what else can we do with these competencies? Is it fixed to workplaces; so, one workplace, one company? Or is it portable?
So, maybe we can consider involving the professional institutions, so that for each level or each competency level, whether it is stackable, micro-credentials, they are all certified and recognised by the industry. And therefore, it is portable for this person throughout his careers.
AI is here to stay. Like most workers, initially, I was, I would not say sceptical, but unwilling to change my style. But now, I think I have got no choice. So, for the past one year, I have been working on it. Even my wife has to use AI to generate videos and pictures, although she is an artist, she likes to draw – but it has become part of her life.
It is a tsunami of change and I support the Motion. As I have previously done in my previous voluntary life as a council member of Institutions of Engineer, Singapore, I am the one who actually proposed working with the Labour Movement to set up the Young Engineers Leadership Programme, including the advanced and the global. I believe that the engineering institutions and the professional institutions could work with the labour unions to certify all such programmes for career pathways for the workers.
Mr Speaker : Ms Eileen Chong.
3.15 pm Ms Eileen Chong Pei Shan (Non-Constituency Member) : Thank you, Mr Speaker. In Mandarin, please.
( In Mandarin ) : [ Please refer to Vernacular Speech .] Mr Speaker, I agree with many of the points raised in the Motion, including the commitment to ensure that no worker is left behind during the AI transition.
However, to realise this vision, we also need to share the gains from adopting AI and ensure that future workers can continue to remain competitive and stay ahead of the curve. At present, when technology is adopted to improve productivity, it is often the employer who benefits. I propose upgrading the Flexible Work Arrangement Guidelines so they are given legislative force, which ensures that employees too, can enjoy AI productivity gains.
When AI improves productivity and work efficiency, we should encourage workers to use the time freed up to be with their families, to rest, to participate in activities and to build relationships with others.
Additionally, the workers of tomorrow – who are the children in our schools today – are also already beginning to encounter AI in the classrooms. Some parents have raised concerns about whether introducing AI to children at Primary 4 might be too early. Neuroscientists have also pointed out that premature or excessive use of AI and technology may make learning too easy, thereby depriving children of the opportunity to develop deep learning capabilities.
Children in their formative years should be learning how to think, ask questions and making judgements. I therefore call on the Government to track and regularly report on the impact that AI adoption in our schools has on the cognitive development of students across different age groups.
In the AI era, independent thinking and judgement are the very skills that AI cannot replace. These are what we should be imparting to the next generation so that they remain competitive and resilient, no matter how the world changes.
( In English ): Mr Speaker, I share the values outlined in the Motion: that growth must be inclusive, that every worker matters and that no one should be left behind in the AI transition. Ensuring that no one is left behind involves more than just ensuring job full growth. It requires a commitment to sharing the rewards of AI driven productivity. It also requires that we ensure that future generations of workers can thrive in an era defined by AI and possibly other technologies have that have not yet been invented.
Mr Speaker, much of the discourse about AI-driven economic transition has been focused on the need for workers to upskill and remain relevant. While these efforts are essential, they do not answer another equally important question: how do we ensure that the equitable distribution of the AI productivity dividend?
Presently, employers are the default beneficiaries. They benefit from either more output from the same headcount or the same output from a lower headcount. Such productivity gains do not automatically become employee gains. Without deliberate policy design, they tend to remain exclusively employer gains.
One way to share the AI productivity dividend with workers could be through time. During the MOM Committee of Supply debate in March, I made a case for flexible work arrangements to be given legislative force. The right to request flexible work is not the same as the right to have it. We should not rely on guidelines that place the burden of action on the employee who can least afford to do so.
I would like to reiterate the call for flexible work legislation today. It has become more salient as we discuss how the AI transition can benefit Singaporeans. As AI generates real productivity gains, the question of whether Singaporean workers will share in these gains as time regained, not just as higher output, is one which the market will not answer on its own. It must be designed for.
What does more time mean in practice? It means the parents who can be present and do more than just paying for tuition classes. It means fewer caregivers having to choose between a job and a family member who needs them. It also means rest, real rest, in a country where 61% of employees feel exhausted and 39% of workers dread going to work. These are not soft outcomes. There are conditions under which human capabilities are replenished and sustained.
Will the Government and tripartite movement commit to prioritising worker well-being alongside employer gains and economic growth as it shapes AI era policy? If so, I urge the Government to begin by giving legislative teeth to flexible work arrangements.
As AI makes company more productive, workers should have a meaningful and enforceable claim on the time that AI frees up. Time to rest and to pursue the kind of human connection that AI cannot replicate, and that our fertility rate is telling us we are running short of.
Mr Speaker, I now want to turn to a group of Singaporeans who matter most to our long-term success. If we say that every worker matters, then we must look at the workers who are not yet in the workforce. And I am talking about our children who are in today's classrooms and some of them in the galleries above.
Right now, our 10-year-olds in Primary 4 are being introduced to AI tools. While there is, of course, teacher supervision and guardrails in place, we should also be asking a more fundamental question. Is this early exposure building their capability and ability to thrive in an AI world, or is it building an early dependence on AI-powered tools? Some parents are already asking. Is Primary 4 too early? What are the real gains? And more importantly, what are the trade-offs?
These are not just parental anxieties. There are also serious questions being asked by neuroscientists. In his book, "The Digital Delusion", neuroscientist Dr Jared Horvath makes a point that should give us pause. When technology makes thinking too easy, the depth of learning disappears. AI is the ultimate offloading tool. It reads, writes and calculates with minimal user input. But our children are not yet experts looking to offload and increase productivity. They are learners and learning requires struggle. It requires the cognitive friction that AI is designed to remove.
If our children start to offload their thinking at age 10, they do not develop the mental muscles needed to spot errors, ask meaningful questions, or form independent views. What we call AI-enabled personalised learning, then risks becoming customised comfort. It feels like progress because it is frictionless, but it may be short circuiting the cognitive development we are trying to support.
This brings me to a concern that I raised earlier today: the equity paradox in AI use. I appreciate Minister Desmond Lee's point about how MOE actively engages parents through Parents Gateway and share guidance on how they can better support their children at home. But not every child in Singapore has a parent at home who is digitally engaged, has the time to act on that guidance and is equipped to scaffold their children's learning beyond what happens in a classroom. Children from less privileged backgrounds with less access to parental guidance and fewer non-screen enrichment activities may end up leaning on AI more heavily, and not less.
For a child who comes home to an overstretched household where there is no one to redirect, question or supervise, AI will always produce an answer, always reduce the friction and always make the thinking easier. That is not empowerment. If AI dependency erodes cognitive development that it is meant to supplement, then the children most at risk are the ones that we are trying our hardest to support.
Mr Speaker, my colleague Assoc Prof Jamus Lim noted during the recent Budget debate that the gap between stronger and weaker university students is no longer about what they submit in written assignment, but about their willingness to question and think beyond the script. These abilities are built or not built over years. If we displace that effort at age 10, our children cannot simply download it at university. And if the children who are doing the most unsupervised AI offloading are those who already have less of these abilities, then we are not closing the gap. We are widening it – earlier.
A global study published this January by the Brookings Institution found that the biggest risk of AI in education is the displacement of effortful thinking during crucial development years. Interestingly, 65% of the students surveyed in this study cited the undermining of cognitive development as the top risk of the use of AI. The children themselves can feel the difference.
Mr Speaker, I am not suggesting that we do not use AI tools in school. I am suggesting that we follow the evidence, not the hype. Generative AI has been public for barely four years. We do not yet have long-term data on how it affects a child's brain. We should not let the speed of a technology cycle outpace the care that our children deserve.
I also appreciate the Minister's update this morning that the Singapore Longitudinal Early Development Study (SG LEADS) by A*STAR will expand to collect data that will help us understand Singapore children's AI usage patterns and how AI usage affects their learning and well-being outcomes. I hope this will also include the impact of using AI tools in education on our children's cognitive development, including their effect on their executive function and skills like critical thinking, reading comprehension and capacity for sustained independent effort. We must ensure that we are preparing our children for a world we cannot predict by giving them the one tool that will always be relevant: a strong and independent mind.
This Motion rightly calls for Singapore's approach to AI enabled growth to be anchored in fairness, resilience and opportunity for all. I agree. And this is why I spoke today about two critical things that the AI transition is challenging us to protect: time and mind.
For the workers of today, we should legislate the right to flexible work so that productivity gains are reclaimed as time for rest, care and connection. And for our children, the workers of the future, we must protect the cognitive friction necessary for learning, ensuring that we are cultivating independent minds that can solve a hard problem without reaching for a digital crutch. The AI transition is not just an economic event. It is perhaps the most significant opportunity in a generation for us to ask what kind of society we will build with the time and capability that technology returns to us. Mr Speaker, I support the Motion.
Mr Speaker : Mr Vikram Nair.
3.26 pm Mr Vikram Nair (Sembawang) : Mr Speaker, I support this Motion. AI is already reshaping our economy. It is improving productivity, enabling new business models and strengthening our global competitiveness. For Singapore, this is an important opportunity. But alongside these benefits, there is also a real concern – how do we ensure our workers are not displaced faster than they can adapt?
Over the course of human history, economic growth has been corelated with the creation of new jobs. When countries moved through industrialisation for example, as industries opened, new jobs were created for people all around. Of course, economic growth can also come not just from increases in labour, but also increases in productivity. And this is also to be welcome because it creates higher paying jobs for those who are working. Singapore has benefited from both of these trends.
Against this backdrop, the conceptual concern with AI is a simple one: will it increase productivity so much that significantly fewer jobs will be needed? The implication for this is that the fewer people who have jobs or alternatively, those that control capital, will get all the benefits from the higher productivity while a large group of people will lose jobs. Essentially, the winners will take a lot more and there will be a larger number of losers.
If we look at developments in other countries, we see that governments are beginning to respond. They do so not by stopping technological change, but by introducing measured safeguards to ensure that workers are treated fairly as the adoption of AI increases.
One area I wish to discuss is the use of AI in the job selection process itself. For instance, in the European Union, the recently adopted Artificial Intelligence Act recognises that AI systems used in employment, such as those involved in hiring, evaluation and performance monitoring, can significantly affect workers' livelihoods. These systems are therefore classified as "high risk" and are subject to requirements such as bias testing, transparency disclosures and human oversight.
This complements Article 22 of the General Data Protection Regulation, under which individuals have the right not to be subject to decisions based solely on automated processing where such decisions produce legal effects concerning him or her or significantly affect him or her. In an employment context, this means that important decisions, such as hiring or dismissal, cannot be made purely by algorithms without meaningful human involvement.
In New York City, the Automated Employment Decision Tools Law requires that AI systems used in hiring or promotion undergo regular bias audits and that applicants are informed when such tools are being used.
While these are not laws which directly prevent job loss, they do ensure that decisions affecting employment are not made with reliance on AI in an opaque or unaccountable manner. They also promote fairness and help guard against unintended discrimination in automated decision-making.
Meanwhile, in countries such as Germany and France, labour laws require employers to follow structured processes before lay-offs, including consulting employee representatives, providing advance notice, and making efforts to retrain or redeploy workers. Our NTUC is engaged in similar activities with our employers here. These requirements may not be specific to AI, but they help ensure that transitions are managed in a structured and responsible manner.
Mr Speaker, these examples suggest that instead of blocking technological change, governments recognise that legislation can provide guardrails and ensure that as companies adopt AI, they do so in a way which takes into account the impact on workers. This is particularly important in maintaining trust between employers and workers. If workers feel that decisions are being made transparently and with safeguards in place, they are more likely to support rather than resist the adoption of new technologies.
For Singapore, we can consider whether there should be clearer expectations around human oversight in employment decisions involving AI. While many employers already adopt such practices, formalising this principle can help ensure consistency across sectors. We can also explore whether workers should have a clearer right to transparency, including the right to know when AI systems are being used to assess their performance or influence decisions about their employment.
In addition, it is worth considering whether we should strengthen expectations around responsible workforce transitions. When major technological changes significantly affect jobs, employers can be encouraged to provide structured support such as retraining opportunities or redeployment pathways.
Ultimately, it is important to recognise that legislation alone is insufficient and must be complemented by strong institutions, proactive employers and workers who are willing to adapt and learn. Our goal should be to maintain a system where businesses remain innovative and competitive, workers feel secure and supported, and opportunities continuing to expand over time.
This leads me to my second point – what we can do for jobs which are most likely to be affected by AI. AI is particularly effective at performing routine and rules-based tasks. As a result, roles in areas such as administrative work and those involving entry-level analysis are more exposed to displacement.
However, the issue is not simply that these jobs may disappear. More importantly, these roles often serve as entry points into the workforce, providing workers with the experience and skills needed to progress. If such opportunities are reduced, workers may find it harder to build careers over time.
In this sense, the risk is not only displacement, but also the gradual erosion of career pathways. Over time, this could lead to a situation where it becomes increasingly difficult for individuals to move from entry-level positions into more skilled and higher-paying roles.
Therefore, beyond retraining in a general sense, our focus should also be on facilitating practical transitions. Workers should be supported in moving into adjacent roles where their existing skills can still be applied and built upon. This makes transitions more feasible, particularly for mid-career workers.
At the same time, companies can be encouraged to redesign jobs so that AI complements rather than replaces human workers. For example, while routine tasks can be automated, human roles requiring judgement, communication and problem-solving can and should be retained and enhanced. This way, AI becomes a tool which increases productivity rather than a substitute for labour.
In practical terms, employers adopting AI can be encouraged to identify adjacent roles for affected workers early on and to provide structured pathways for redeployment into these roles. This may include short, focused training modules or job redesign which allow workers to gradually build new competencies.
Such measures help ensure that workers are not abruptly displaced, but are instead guided through a managed transition where their skills and roles within their organisations are preserved. The unions, the Government and employers can work together on a framework for this.
In relation to the creative industries, including music, writing and acting, we should consider if legislation or further protection is needed in relation to copyrighted material being used to train AI and whether remedies for this should be purely private or whether there is scope for the Government to provide a framework for such materials to be protected. This may include the right to one's image and voice. If left purely to private law, only the well resourced would be able to take up the matter, whereas if there is a framework for this, individual artistes, writers and others might be able to benefit from such protection.
Singapore has navigated many economic transitions successfully over the years. Each time, we have combined openness to change with a commitment to social mobility and shared progress.
The transition to an AI-driven economy will be another such test. It will require us to strike a careful balance between innovation and protection. If we approach this thoughtfully, we can ensure that AI becomes a source of opportunity and that growth remains inclusive. I support this Motion.
Mr Speaker : Mr Fadli Fawzi.
3.35 pm Mr Fadli Fawzi (Aljunied) : Mr Speaker, today's Motion rightly recognises the transformative power of AI and affirms that AI-enabled economic growth must remain inclusive.
My speech consists of three broad points. First, we must protect the economic positions of workers and prevent the economic fruits of AI from accruing solely to those who own the AI models or produce the hardware powering AI. Second, we must ensure that AI-resilient employment pathways remain viable for Singaporeans. Third and most importantly, we must hold firm to the idea that technology should serve humanity and not the other way around, because ultimately, the goal is not just growth without joblessness. It is growth without losing who we are as humans and as Singaporeans.
Sir, the second limb of the Motion statement calls upon the House to emphasise that Singapore's approach to AI-enabled growth must be anchored in fairness, resilience and opportunity for all, while the fourth limb asks the House to affirm that economic progress must remain inclusive and that Singapore must not have jobless growth.
These are extremely important goals, because if the rise of AI is not managed properly, it could represent not just technological disruption but a recalibration of power between labour and capital. For example, last month, Meta announced that the keystrokes and workflows of every one of its employees in the US will be recorded. Screenshots will be taken occasionally throughout the workday. All this data will form datasets used to train AI systems that could one day replace these employees. Other companies could follow suit soon.
While this is being done in the US, we in Singapore should ask whether companies such as Meta should be allowed to harvest employee data in this way without any clear safeguards. Should there not be stronger protections about how such data is collected and used? Should workers not have a stake in the value that is created in their own data?
If we fail to address these questions, we risk sleepwalking into a future where wealth and power are concentrated in the hands of a few technology giants who increasingly use AI to take over work previously done by their human employees. These firms will continue to be the engine of economic growth, investing in larger and larger data centres and more powerful semiconductor chips that continue to generate gross domestic product growth.
As a country, Singapore may benefit through our shareholdings in and partnerships with these tech companies. However, we must ensure that these benefits, which primarily accrue to the capital owners, do not come at the expense of the labour force. These developments pose a real risk that workers who are replaced may see their economic power steadily eroded, with a larger number of workers having to chase after a smaller pool of lower-paying jobs.
I do not mean to be alarmist in suggesting that if unchecked, what could emerge is a form of digital serfdom, a system where workers like serfs of old are not bound by land or feudal lords but by algorithms. We are already seeing the beginnings of this future play out in real time. Already, platform workers work in service to black box algorithms that have a great deal of control over their earnings and how many hours they work. As AI advances, many cognitive and white-collar jobs will become increasingly automated and possibly, under the control of algorithms. Roles once considered secure may no longer be so.
As such, we need to strengthen frameworks for worker protection in areas such as retention benefits and the rights of workers over data they create at the workplace. AI-enabled growth must not come at the cost of workers and a further tilting of the balance of economic power towards capital owners.
Sir, the third limb of the Motion asks the House to equip and support workers and enterprises to seize new opportunities and advance together. Even as AI threatens to automate and replacement many existing roles, there remain many forms of labour that are difficult to automate using AI.
For example, plumbers, electricians, air-conditioning technicians, phlebotomists and other skilled trades have been assessed to be much less likely to be replaced by AI. These jobs are essential pillars of a functioning society. Yet for too long, we in Singapore have undervalued these roles, both economically and socially. If we are serious about ensuring that growth remains inclusive, we must correct this imbalance. In many other first world societies, the job of a plumber, garbage collector or an air-conditioning technician pay high enough to allow a middle-class lifestyle. This is not the case in Singapore.
We have made a policy choice to fill these roles with lower-paid foreign workers while our local workers are channelled into high-paying white-collar jobs. While this has worked well for us for decades, this may no longer be sustainable as generative AI threatens to reduce the number of well-paying white-collar cognitive roles.
We must therefore raise wages and improve career pathways in blue-collar sectors that are currently less attractive and yet also less vulnerable to displacement by AI. We must elevate their status not just through policy, but through culture and education so that Singaporeans will no longer see such jobs as undesirable.
This may require difficult trade-offs. For example, should we recalibrate our policies in certain sectors to ensure that wages for local skilled trades rise meaningfully and attract more Singaporeans to fill these roles? At the same time, we must make better use of our strong vocational institutions. Our ITEs and polytechnics should guide more students towards specialised high-value trades. In an AI-driven future, the dignity of work must not be tied solely to whether a job is white-collar or high tech. We must expand the range of jobs that Singaporeans considered attractive and meaningful.
Mr Speaker, AI will undeniably determine the future of our economy, our society and our lives. But we should not allow AI to come to define us as humans, as citizens and as Singaporeans. I say this because the question before us is not merely whether AI will create or destroy jobs. The deeper question is this: what kind of society and what kind of human beings will we become in an age shaped by AI? Because Sir, if we are not careful, we may succeed economically, yet diminish ourselves in more fundamental ways.
For example, AI has created a world where knowledge is no longer scarce. Texts can be summarised, essays can be written and equations can be solved in seconds. I spoke before about my own experience as an undergraduate struggling through dense text. It was slow, often frustrating work, but it was through that struggle that I learned how to think, how to question and how to make sense of the world.
Today, however, the effort required to complete any cognitive activity has collapsed to an extent far greater than when the calculator replaced the abacus or the typewriter replaced the pen. In encouraging our students to leverage AI, how can we ensure that they can continue to learn how to grapple with ideas, how to formulate arguments, how to problem solve and how to cultivate intellectual independence?
Sir, I reiterate my caution – that Singapore must become an AI-resilient society and not an AI-reliant one. By constantly outsourcing our tasks to AI, we may erode or undermine our capacity for creativity, imagination, judgement and even empathy, or let these practical skills atrophy from lack of use.
The danger here is the temptation to use AI as a shortcut for thinking through and solving problems. On the one end, there are those who regard AI as a kind of second brain – outsourcing memory, decision-making and even aspects of judgement to ChatGPT or Claude.
It is true that AI tools can sharpen our thinking and serve as intellectual aids. However, in creating a layer of artificial mediation between us and the world, I am concerned that AI would dull our capacity to make sense of the world. By making sense of the world, I mean the ability to interpret, comprehend and coherently perceive the world around us on our own terms, through our own cognitive efforts. And over time, this would involve reflective trial and error, balancing our considered interpretations and judgements of the world with how the world comes to bear upon us.
Interpretation and judgement are practical skills that must be honed through constant and regular use. And we develop these skills by exercising, testing and challenging them. Making judgements about the world and what should be done is a distinctively human task that should not be easily surrendered.
I am not against the idea of a second brain. My worry is more specific – that the reliance on a second brain, if left unchecked, will weaken the equity and reflexes of the first brain.
On the other end, we see people forming emotional attachments with AI companions. These are relationships that stimulate empathy but do not truly reciprocate it. It demonstrates the real risk that people can lose sight of human relations in the real world. If people see these AI companions as a comforting shortcut to finding companionship in contrast to the hard work of developing friendships with others around us, we may see a further impoverishment of our social networks.
In both cases, the danger is the same. We begin to substitute authentic human experience, sense-making and judgement with artificial approximations. And when that happens, we may gradually lose our ability to navigate the world with clarity on our own terms. Mr Speaker, in Malay.
( In Malay ) : [ Please refer to Vernacular Speech .] Sir, a final point. There is a traditional Malay art form that is close to my heart: the pantun – a poetic form with its own specific rules. Made up of four lines, the pantun has a specific metre and rhyme scheme. Its imagery is typically drawn from nature and scenes of everyday life, to communicate important social values and advice.
A pantun that does not follow these structures and conventions is usually not regarded as a good pantun, if it may be called one at all. Many pantuns are still known among Malays by heart, passed down orally through generations. In short, the pantun embodies a tradition – connecting the Malays today to our forefathers before.
Hence, I want to ask: do we lose something valuable if we teach students to use AI to generate pantuns, rather than discovering the fun of experimenting with the lines themselves? Does the skill of using AI to generate pantuns necessarily translate into the craft of writing a good pantun, or even the aesthetic sensibility to appreciate the art form? And what is the long-term impact to the Malay language, culture and tradition when a cultural motif is reduced to an AI output?
( In English ): Sir, a final point. There is a traditional Malay art form that is close to my heart – the pantun. This is a poetic form with its own specific rules. Made up of four lines, a pantun has a specific meter and rhyme scheme. Imagery is usually drawn from nature and scenes of everyday life, to communicate important social values and advice. A pantun that does not follow these structures and conventions is usually not regarded as a good pantun, if it may be called one at all.
Many pantuns are still known among Malays by heart, passed down orally through generations. In short, the pantun embodies a tradition, connecting the Malays today to our forefathers before.
Hence, I want to ask: do we lose something valuable if we teach students to use AI to generate pantuns rather than the fun of experimenting with the lines themselves? More importantly, does the skill of using AI to generate pantuns necessarily translate into the craft of writing a good pantun or even the aesthetic sensibility to appreciate the art form? And what is the long-term impact to the Malay language, culture and tradition when a cultural motif is reduced to an AI output?
Sir, I am not suggesting that we return to a time before AI use. We have to adapt. But we need discernment. We must be clear-sighted about what AI can and cannot offer and always ask what is the purpose AI is serving and whether it is fit for that purpose? We must avoid being boxed in into an AI-centric gaze in which we are left with a narrow and artificially mediated understanding of reality and an impoverished capacity to make sense of and relate to the world and those around us. Sir, let me close my speech with a pantun.
( In Malay ) : [ Please refer to Vernacular Speech .] When delivering a pantun in Parliament;
It is best not to use AI;
The Malays are cultured and courteous;
The poet's inspiration will not be abandoned.
( In English ): When delivering a pantun in Parliament, it is best not to use AI. The Malays are cultured and courteous, the poets' inspiration will not be abandoned.
3.51 pm Mr Speaker : Order, we have been in the Chambers for close to five hours. I propose to take a break now. I suspend the Sitting and will take the Chair at 4.10 pm. Order, order.
Sitting accordingly suspended
at 3.51 pm until 4.10 pm.
Sitting resumed at 4.10 pm.
[Deputy Speaker (Mr Xie Yao Quan) in the Chair]
An Artificial Intelligence (AI) Transition with No Jobless Growth (Motion)
[(proc text) Debate resumed. (proc text)]
Mr Deputy Speaker : Mr Kenneth Tiong.
4.10 pm Mr Kenneth Tiong Boon Kiat (Aljunied) : Deputy Speaker, I declare my interest as a director of a company that makes AI-enabled applications and consults on the same.
In the three and a half years since ChatGPT's release, I have had two moments of awe and dread. The first was in November 2022. GPT-3.5 could iterate on software features, generate ideas, write code. Five years ago, it was received wisdom that everyone should learn to code. Today, coding ability is cheap and abundant. Computer science graduates, even from top schools, like Stanford, are finding it difficult to find jobs. GPT-2 was a toy that generated amusing limericks. Three years later, its successors have made an entire profession's scarcity disappear. We used to talk about prompt engineering in 2023 and 2024. That talk has died down too.
The second was in November 2025, when Anthropic released Claude Code, a reliable AI agent paired with a frontier model. I could leave the computer running overnight and there would be work done at the end. It is a different experience from chatting with a chatbot. The chatbot engages you in back and forth, refining your ideas, indulging your whims, red teaming your speeches. The agent, unless it needs clarifications, just goes and does things. It may be off by a bit, but you give your input and it takes another five or 50 minutes before it comes back with the problem solved. A very smart junior colleague.
Now we have AI agents, Claude Code, Codex tools, that have made me, if I may borrow Internet lingo, Claude-pilled. I use Claude Code for my own work. I can give it the most wishy-washy specifications, and it returns the most wonderful data workflow or website layout. For someone who could never build a pretty website to save his life, it is liberating.
There is a spirit of play in working with these tools that every Singaporean deserves to experience. It is an exhausting world it heralds – software engineers pulling 80-hour weeks while running multiple AI agents overnight so that someone, human or machine, is always on the clock. Jobseekers, especially recent graduates in white-collar work, applying to hundreds of jobs without a single interview. Job portals, like LinkedIn, have become memory holds for resumes, where the lived experience is like shooting an application into the void.
The pace of change humbles us all. I am suspicious of any assertion that starts with "AI will never", because the shelf life of those predictions tend to be measured in months. What concerns me is not the destination, but who gets left behind on the way there and whether we are building the institutions to ensure no one does.
I have three propositions. First, that access to premium AI, especially AI agents, must be universal, not gated by course enrolment or union membership. Second, that we must treat the handful of companies building frontier AI with the same strategic seriousness we bring to bilateral relations with countries, because their decisions on pricing, access and deployment now shape our productivity frontier as directly as any trade agreement. Third, that we must buy time for workers by upgrading our retrenchment framework for AI-speed displacement.
Sir, I believe access to premium AI, especially AI agents, is a right, not a privilege. Intelligence, in the sense of uplift, should not stratify according to wealth. I spend a couple of hundred dollars a month on these tools because they are game changing. But for those who cannot afford to, it bakes in inequality from the start. Does it simply disqualify them from the off?
The Government has partially adopted the 2024 suggestions of my colleague Gerald Giam to provide universal premium AI model access. The SkillsFuture Premium AI Access Scheme, six months of free tools for Singaporeans who enrol in selected courses, is a step in the right direction. Likewise, NTUC's subsidies covering 21 AI tools. These are good starts, but they are unnecessarily gated behind course enrolment and union membership. And critically, they likely will not cover AI agents – the tier where the real productivity gap will open.
Why does this matter? AI agents are expensive to run. We may hope agent access follows the cost curve of Internet bandwidth or compute, but there is no necessary reason it should. It is an empirical question.
Anthropic's CEO said in January that 80% of its revenue comes from enterprise customers, driven by API calls on a pay-per-token model. If agents remain enterprise-grade by default, then individual citizens – jobseekers, freelancers, retirees – are locked out of the tier where the real productivity gains are being made.
Three possible directions.
One, negotiate sovereign access – a bulk licensing agreement with frontier AI providers for volume-discounted AI agent access for all citizens.
Two, if agent access is employer-provisioned in the market, make it universally so, require companies above a certain size to provide agent-grade AI to all employees, the same way we require CPF.
Three, if frontier agents remain too costly, identify a minimal, viable agentic tier and fund that universally.
Will the Government make premium AI access a universal entitlement, rather than gate it behind course completion or union membership?
Sir, I learnt recently that even AI engineers at the top two or three frontier AI labs are worried about falling behind because they cannot use Claude Code. And having just returned from China, I learned first-hand that one cannot use Claude Code there at all. Anthropic blocks API calls from mainland China, Hong Kong and Macau entirely.
If even engineers building frontier AI are desperate for access to one another's tools and if entire countries can be locked out, then access is not a convenience. It is a strategic capability. And the question for our country is whether we will secure it or whether we will be price-takers forever.
There are, perhaps, three to five companies in the world whose decisions on pricing, access and deployment will shape every economy's AI trajectory. When Anthropic or OpenAI decides what to charge for agent-tier access or whom to serve, that decision shapes Singapore's productivity frontier as directly as any trade agreement.
We should therefore treat this class of companies – frontier AI firms that have crossed a threshold of systemic importance – with the same strategic seriousness that we bring to bilateral relations with countries. Not because they are sovereign, they lack the durability and legitimacy of states and remain subject to home-state law, but because their decisions carry sovereign-grade consequences for our economy and we should engage them accordingly.
What does that mean in practice? Four things.
First, negotiate access at the sovereign level. In the possible future where frontier AI agent costs go up, not down, Singapore should seek bulk licensing agreements for agent-tier access the same way we negotiate energy supply. This means accepting that frontier AI access may be a permanently higher line item in the national expenditure and procuring it systematically, because the alternative – citizens priced out of the tools that define productivity – is worse.
Second, we trade based on what we have. Nvidia CEO Jensen Huang has described the AI stack as a five-layer cake: energy, chips, infrastructure, models and applications. In my view, we do not have energy at scale. We do not have frontier model capability. At the application layer, there is little moat outside of the knowledge agglomerations we can build for ourselves. We would be competing with some of the highest cost bases in the world.
But Singapore Inc builds good data centres. And we are among the world's leaders in water reuse and integrated water management, which is a binding constraint on data centre expansion across water-scarce regions in Southeast Asia. If we position ourselves as the infrastructure partner of choice for this region, that is real leverage – something we can bring to the table in exchange for access, for pricing and for presence.
So, when a company like Anthropic or OpenAI approaches us, we should be their preferred regional bilateral partner in rolling out and scaling their data centre build-up regionally, as well as all the infrastructure needed to make these data centres work.
Third, we need to attract real technological presence. We should seek frontier AI companies establishing development offices here – not predominantly sales offices, which was the experience with the FAANG companies in the 2010s. And I would prefer that we be quality-conscious. Most companies that call themselves "AI companies" are not frontier AI companies. We need targeted strategy and engagement with frontier AI companies specifically.
Fourth, get Singaporeans inside these labs. Once you are in the frontier AI ecosystem, it becomes much easier to circulate within that group of companies. I would welcome the Government doing some fact-finding – engaging our local and overseas Singaporeans already in these companies and roles, understanding how they or their colleagues got hired and disseminating that to our students and technical researchers. Right now, anecdotally, half a million to million-US-dollar salaries, excluding equity, in the US for AI researchers are fairly common. So, it is clearly in our interest to figure out how to get more Singaporeans into this tight labour market. What I would really like to see is a Skills Framework for frontier AI lab researcher.
Sir, my last point is about the transition. Let me start with a person. In Hangzhou, a quality assurance supervisor named Zhou joined a tech company in late 2022 at RMB25,000 a month, about S$4,800, reviewing AI model outputs for accuracy and safety. In 2025, his employer decided an AI model could do his job. They offered him a reassignment at roughly 40% less pay. He refused. They terminated him. Zhou went to arbitration and won. The company sued and lost. The company appealed and lost again, at the Hangzhou Intermediate People's Court. The ruling was published on 28 April this year, three days before International Workers' Day.
The court's reasoning is worth our attention. The company argued that AI had made Zhou's role obsolete – a "major change in objective circumstances", justifying dismissal under China's Labor Contract Law. The court disagreed. AI adoption, it held, is a deliberate business strategy, not an unforeseeable event. A company that chooses to automate cannot unilaterally shift the full cost of that decision onto the worker. The company had not shown the contract was impossible to perform and the reassignment at 40% less pay was not a reasonable alternative. The court added that companies should prioritise retraining workers and helping them transition to higher-level roles.
The principle, that a deliberate business decision should not externalise its full cost onto the worker, deserves serious consideration in Singapore.
If a Singaporean Zhou were retrenched tomorrow under our existing framework, would he win? Our existing Tripartite Advisory on Managing Excess Manpower and Responsible Retrenchment (TAMEM) is advisory, not statutory. An employer can lawfully automate a role and terminate the worker without first attempting to redeploy or reskill them. And the public purse, through SkillsFuture and Workforce Singapore, would pick up the cost of that worker's transition. There is no AI-specific notice period. There is no statutory redeployment-first obligation. There is no individual cause of action for the worker to challenge the reason for his or her termination.
The data suggests we are entering the zone where this matters. MOM's own fourth quarter of 2025 Labour Market Report reported about 14,490 retrenchments in 2025, up from about 13,000 the year before. PMET retrenchment incidence reached 10.1 per 1,000 resident employees, above the pre-recessionary norm of 8.0 set during 2015 to 2019. Retrenchments were concentrated in Financial Services, Information and Communications, and Professional Services – the most AI-exposed sectors. Information and Communications employment declined outright in 2025.
The Government announced the Tripartite Jobs Council on 30 April. I, of course, welcome its intent. But it creates no new powers, no new obligations on employers and no new rights for displaced workers. How does the Government intend for this Council to work?
Too often, what workers experience is not a frank conversation about AI-driven restructuring but a Performance Improvement Plan, a process that, in many cases, is a bit of "wayang", designed to paper over a predetermined outcome. I foresee that such potential misleading reasons may be given and workers must have the power to be able to challenge this.
I propose three directions. First, a 90-day mandatory transition notice before AI-driven role elimination. Second, a re-deployment-first obligation, retraining or reassignment before AI-driven termination. These provisions will slow the velocity of AI disruption; and velocity is what determines whether adjustment is possible. Third, for workers to be able to substantively challenge the reasons for their termination if they feel they are misleading, so that these AI restructuring protections will be real.
Sir, in closing. Finland gave people unconditional cash as income. They were happier, less stressed – and the great majority still walked into the employment office and asked for work. The American pollster David Shor polled Americans this year: three to one, across every political persuasion, they chose job creation over direct transfers. People, when offered the choice between a universal basic income and employment, invariably choose employment. Not because they are irrational, but because a job is where you are needed, and being needed is not something a universal basic income can replace.
So, no jobless growth – yes. But more than that: no growth where the gains are captured disproportionately by capital and the burden of adjustment falls on labour. Universal access, so intelligence is not rationed by wealth. Strategic engagement, so we are not price-takers in our own future. And a retrenchment framework where the company that decides to automate bears the cost of that decision before the worker does.
I do not think the awe and dread goes away. But in a country that builds for its workers, there is hope for a brighter future. Thank you.
Mr Deputy Speaker : Mr Sanjeev Tiwari.
4.25 pm Mr Sanjeev Kumar Tiwari (Nominated Member) : Mr Deputy Speaker, before I start, my greetings to my fellow unionists on both sides of the viewing Chamber. Thank you for the support.
Mr Deputy Speaker, as unionists, our role is not just to support change, but to ensure that the change works for our workers. With all the discussions so far, there is no doubt that AI brings about new opportunities, new tools, new ways of working and the potential for better jobs.
For the Labour Movement, an AI transition with no jobless growth must means three things. First, AI augments workers, it does not replace them wholesale. Two, the productivity gains from AI are reinvested into people, into training, into new industries, into better wages and real growth. Three, no worker is left to navigate this transition alone. I will speak to each of these in turn.
Ensuring productivity gains are shared. Mr Deputy Speaker, if we are not deliberate, the gains from AI will not automatically be shared. They will tend to concentrate in firms with the scale and capability to deploy it. The business case for transformation is compelling and I support the Government's announced efforts to support enterprises on this to accelerate their AI adoption, so that they can seize new opportunities.
However, I call on businesses seeking to grow their pie, to be laser-focused on job redesign and training to bring their workers along with them. Here, I must appreciate the many hon Members who have supported this call.
Globally, there are warnings that AI could significantly reduce entry-level white-collar roles in the coming years. Closer to home, DBS Bank has announced plans to reduce its contract and temporary staff by around 4,000 across various markets, with AI adoption.
While firms rationalise their workforce sizes and skills mix, we must be clear-minded that not all of such rationalisation may translate into economic growth that enables our families to thrive, children to flourish and seniors to enjoy their golden years.
Workers are therefore paying attention. They want to know that there are pathways for the gains from AI to benefit workers, and not just management and shareholders. Many workplaces are only at the beginning of figuring out what such pathways need to look like and the pitfalls of failing to provide such pathways.
As mentioned by some Members, recently, courts in China have been active in reviewing cases regarding the dismissal of employees due to AI-related restructuring and choosing to protect labour rights against unfair AI-related lay-offs. In one case, the arbitration panel clarified that AI replacement was not valid grounds for dismissal. In another case, a massive role and salary reduction due to AI taking over the work was not considered a reasonable re-assignment proposal.
I would like to believe that we will not see such court cases in Singapore. Hence, companies must be held to a human transition standard. When a company deploys AI that eliminates roles, it should be required to have a transition plan, re-deployment offers, funded retraining, phased timelines.
It must work with the unions through the CTCs or the tripartite frameworks, to ensure this is managed together. We should make this a baseline, and not the exception. The social contract between employer and employee must evolve alongside technology.
Unions should be involved early in the company's transition plans to integrate new technology, implement job redesign and transition workers to handle new work. In roles that AI is not displacing, AI is increasing the speed, density and complexity of work rather than reducing it.
However, I caution that when AI has pushed human boundaries and job redesign is not done adequately in tandem, there is concern that the working environment will be unsustainable, with very high intensity and pressures that can lead to burn-out, fatigue and poorer psychosocial health. We already see this in many of today's workplaces, especially for PMEs, where work is increasingly outcome-based and the boundaries between work and personal life are more blurred. AI risks accelerating this trend even further.
The psychosocial implications of AI enablement deserve further attention too. Our time-based employment protections were designed for a less digitally connected era with more fixed working hours and clearer boundaries. Today, these new emerging work patterns suggest there is room to further evolve how we can support workers.
This is where our unions in Singapore play a critical role. Through collective agreements, company-level engagements and multi-company initiatives, such as the Queen Bee partnerships for the NTUC's CTC initiative, unions ensure that productivity gains are translated into better jobs, better wages and most importantly better working conditions, not just higher output and shareholder returns. In addition, when AI is deployed to support hiring decisions and performance reviews, we must watch for unintended biases and ensure safeguards for the confidentiality of information.
Supporting workers through transitions. Even when gains are shared, there is a harder question we must confront: what happens to workers whose jobs are displaced altogether?
Mr Deputy Speaker, the standard response we often give is to reskill, adapt, move on. That advice assumes that workers have time, the financial buffer and margin for error to take such risks. However, not all workers have this luxury. I want to be honest, because vague optimism is a luxury that the displaced cannot afford and we must make sure the system watches for this.
This is especially so for our mid-career and older PMEs. They are workers with mortgages and bills to pay, with children still in school and often, elderly parents to take care of. They are not just managing careers. They are carrying entire households.
For them, transitions are high stakes. A failed transition is not just a momentary setback. It can mean prolonged unemployment, income loss and long-term repercussions for their loved ones. And this we are already seeing in the data that is provided by MOM, where there has been a rise in PMET retrenchments compared to the pre-recession norms in 2019, reflecting their greater exposure to sectors undergoing restructuring.
This is where we must go further to support our breadwinners and their families, ensuring that reskilling leads to real job outcomes, that transitions are supported and that pathways to good jobs are clear, especially for those making mid-career shifts.
Instead of welfare, we have workfare. And instead of minimum wages, progressive wage models for key segments of our workforce. For the AI era, we can seek better support for our PME jobseekers. We must continue close monitoring of those who applied for jobseeker support and help them bounce back as soon as possible for the next better job.
Mr Deputy Speaker, giving workers a genuine voice in AI adoption, where the first two pillars of sharing gains and supporting transitions cannot happen without the third. Giving workers a genuine voice in how AI enters their workforces is equally important.
Across advanced economies, one principle is becoming clear, worker voice must be part of how technology reshapes work. In countries, like Belgium, unions and employers are already working together to establish norms around after-hours communication, workload and staffing.
In Singapore, we have a strong foundation in our tripartite model. But as a unionist, I want to make a broader point. If we are serious about ensuring AI is used fairly and that the gains from AI-driven transformation reach workers, then we must welcome unions to represent PMEs who are most vulnerable in the AI era.
PMEs are not a monolithic group – an engineer in aerospace, a financial analyst in banking, a project manager in the tech phase – very different work environments and go through very different impacts of AI transformations. Unions are able to shape workplace norms from ground up, in a targeted manner that recognises diversity. Hence, employers should consider allowing unions to represent PMEs.
The mechanism for having the workers voice at the table is already here – it is the CTC. Through the CTC, unions work directly with management to chart out transformation roadmaps, redesign jobs and upskill workers so that no one is left behind when a company transforms.
Let me just give one example. SBS Transit, with support from the National Transport Workers Union and the CTC Grant, overhauled its bus maintenance operations using AI. The company implemented AI-powered diagnostic systems for predictive maintenance and instead of cutting jobs, it created a new diagnostic expert career scheme for over 50 workers. Such examples must be amplified and more employers should do such things.
We must continue to leverage this to support workers and enterprises in the AI transition. More recently, NTUC is partnering global technology leaders, like Amazon Web Services (AWS) and Huawei, to equip 100,000 workers and 100 enterprises with AI skills through the CTC ecosystem.
At the individual level, union members can also tap on union support for up to half of the subscription fees for AI models. This is in addition to the six-month subscription provided through the Government. I hope more leading multinational corporations can work with our unions to provide more training and uplift the AI skills of workers for our collective future.
In conclusion, Mr Deputy Speaker, I call for Singapore to move forward in these areas.
First, on sharing gains, we must ensure that as enterprises transform with AI, the gains are shared with workers through fair wages, better working conditions and clearer norms around work intensity, including expectations on after working hours communication and responsiveness. They must be calibrated to our Singapore context but clear in intent.
Second, in supporting transitions, we must strengthen how we support workers through the AI transition by ensuring that AI enables enterprises to unlock new growth, that training leads to real job opportunities, that reskilling is twinned with job redesign and that workers are not left to navigate these changes on their own.
Third, in giving workers voice, workers and their unions must be engaged early when AI is introduced into workplaces. Not just on hiring and other employment decisions, but on how AI changes job scopes, workflows and performance expectations. This means welcoming unions' ability to represent PMEs, and scaling mechanisms, like the CTC, so that worker voice is embedded in every transformation journey.
An AI transition with no jobless growth is not a slogan. It is a commitment. A commitment that growth should mean something to everyone, not just those at the top of the economic pyramid. But to the nurse, the teacher, the logistics worker, the small business owner, the young person entering the workforce for the very first time. They are not footnotes in the story of technological progress. They are the reason progress should matter at all. These are not competing priorities; they go hand-in-hand and we are standing at one of those rare inflection points in history where choices we make today will echo for the next generations.
That is where the tripartite partners must deliver, must make this happen and that is also what I hope this House will help us deliver. I strongly support the Motion. [ Applause. ]
Mr Deputy Speaker : Mr Sharael Taha.
4.38 pm Mr Sharael Taha (Pasir Ris-Changi) : Thank you, Mr Deputy Speaker. Mr Deputy Speaker, Sir, I declare my interest as someone working in the aerospace and advanced manufacturing industry, focusing on strategy, digital transformation and AI transformation.
Mr Deputy Speaker, I stand here today with a deep sense of gratitude. I grew up in a very ordinary Singaporean home with working class parents, but was given an extraordinary opportunity to build and retrofit advanced factories across the world.
From Germany to the United Kingdom, from North America to Asia, I have had the privilege of building and working on industry 4.0 facilities, cutting edge factories equipped with some of the most sophisticated machines ever created, machines that produce components of such precision that they can only be manufactured in a handful of places in the world and facilities that assembled some of the most advanced engineering systems ever built.
Mr Deputy Speaker, Sir, after all these years, one lesson stands above all else. It is not the machine, nor the technology that determines success. It is the people. I have seen factories with the best technology money can buy struggle because of misalignment, because of distrust, because workers, unions, management and government were not moving in the same direction. And I have seen more modest facilities outperform expectations because everyone was aligned to a common purpose.
That is why this debate matters, because when we talk about transformation, especially one driven by AI, we are not just talking about technology. We are talking about people, about workers, about livelihoods, about dignity.
And I want to acknowledge our Labour Movement, NTUC Secretary-General, Mr Ng Chee Meng, and Ms Yeo Wan Ling and Members of this House, Mr Mark Lee and Mr Saktiandi Supaat, for putting forth this Motion. Their position is not just right. It is also timely.
It is timely because it complements the direction set out by Prime Minister Mr Lawrence Wong in his Budget speech and the May Day Rally, where he spoke about how Singapore must confront the realities of AI and global uncertainty while standing firmly with our workers. And even as technology reshapes our economy, we must not leave our people behind.
Our Labour Movement has echoed this with clarity and conviction. That this transition must drive Singapore's next phase of growth, but it must be anchored in fairness and opportunity for all. We must equip both workers and enterprises to seize new opportunities so that progress is not just created but shared. This is the promise we must make to our next generation.
Mr Deputy Speaker, having worked across different countries, I have encountered many labour movements, many focus on protecting jobs, protecting specific jobs that exist today. And often to protect the jobs of today, they invariably have to resist change, even when the tide of change is inevitable. But what we have in Singapore is different.
From my enterprise experience, our tripartite partnership between the NTUC, Singapore National Employees Federation and the Government is unique, so unique that many of my overseas colleagues are genuinely puzzled by how well it works with real positive outcomes for all, because our unions do not just protect jobs, they protect workers.
And our union leaders, like brother Samad from the Union of Power and Gas Employees, brother Fahmi from United Workers of Electronics and Electrical Industries, brother Poobalan and Goviden from SATS Workers' Union, brother Gabriel from the Amalgamated Union of Statutory Board Employees and many of our union leaders here, stand with the workers, not just for where they are today, but for where they need to be tomorrow. They focus on keeping workers relevant, employable and ready to take on better opportunities as the economy evolves.
And that, Mr Deputy Speaker, is something that we should never take for granted. Allow me to frame my position on the motion around three key ideas: transformative power, opportunity for all and jobless growth.
First, on the transformative power of AI. The impact of AI can be understood at three levels, the individual, the enterprise and the industry. And at every level, success depends on how well workers, businesses and Government work together. At the level of the individual, AI is a force multiplier.
It enhances productivity, augment skills and allows each worker to do more, do better and do faster. Many of us are already experiencing this today through tools, like Open AI, ChatGPT, Microsoft Copilot, Claude, Gemini and Canva, but this transformation does not happen by chance. Workers must be prepared to learn, adapt and continuously upgrade themselves, and businesses must invest in training, redesign jobs and empower their workforce to use these tools effectively.
And the Government must provide the support structure, strong skills framework, accessible training pathways and broad-based access to technology. That is why I am encouraged by the support the Government is providing to help Singaporeans adopt AI tools, further strengthened by NTUC's initiative to subsidise AI subscription to its members. This is important because AI cannot become a tool only for the privileged few. It must remain accessible to the masses so that every worker has the opportunity to improve productivity, strengthen capabilities and participate meaningfully in Singapore's next phase of growth.
At the level of the enterprise, AI enables better decision, sharper operations and greater efficiency. It turns data into insights and insights into action. But to realise this, companies need the right framework. Workers need the right capabilities and the Government must provide the right environment to scale transformation responsibly.
At the level of the industry, AI creates entirely new value. It reshapes business model. It transforms competition and unlocks new growth. In a tight labour market, like Singapore, this enterprise and industry transformation must go hand in hand with deep business process re-engineering and meaningful job redesign, as mentioned by a few of the Members here.
And speaking from my own experience in the industry, the decision to adopt AI is really just about tax incentives, as characterised by Member Mr Andre Low. The real driver is how we upskill and reskill our workers so that they can take on higher value-added jobs that are increasing in demand and the opportunity cost of not doing so.
The tax incentives help companies invest in the necessary AI tools and infrastructure. But equally important are the schemes that support workers through this transition, whether it is programmes such as the SkillsFuture Workforce Development Grant, NTUC CTC Grant, the Union Training Assistance Programme, and Workforce Singapore's Career Conversion Programmes, these initiatives support job redesign, training and even provide wage support while workers undergo upskilling and reskilling.
Very often, AI adoption is not a binary choice between technology and workers. It is about re-engineering business processes so that enterprises can do more with a more capable, more skilled and more productive workforce to take on the challenges of the world.
These values, unlocked by AI, must be shared – shared so that workers are uplifted, businesses grow stronger and society moves forward together. But ultimately, this is a social compact we must continue to uphold between workers, businesses and government – one anchored on fairness, inclusion and shared progress.
Taken together, AI is not just any other technology tool. It is a system-wide transformation. That is why tripartism matters even more in the age of AI. And I am heartened to hear the position of the Labour Movement.
Second, on opportunity for all. Mr Speaker, this transition will create economic growth. But it is also a period of real uncertainty. We must acknowledge these challenges. Our fresh graduates are feeling it. Many struggle to secure permanent roles. Multiple internships are becoming the norm. At what point do we ask whether this becomes a substitute for proper employment?
As AI reshapes professions from developers to lawyers, how do we ensure there are still meaningful entry points for our young people? If companies begin to question entry-level roles, then we must also ask, who will train the next generation of our workforce?
Our mid-career workers feel this even more deeply. With families and responsibilities, they worry that job transformation may render their experience less relevant. Our blue-collar workers – our technicians, our operators and drivers – are asking hard questions about automation and the future of their jobs.
These fears are real. If left unaddressed, they can divide our society. That is why this opportunity for all cannot be left to market forces alone, a point also raised by Member Poh Li San. It requires companies and especially middle managers to give fresh graduates a real chance. It requires businesses to redesign jobs and invest honestly in upskilling and reskilling. It requires all of us to work together so that every Singaporean can find their place in this new economy. Only then can we say that this is not just growth, but opportunity for all.
Finally, on jobless growth. In many economies, jobless growth means growth without jobs. But in Singapore, our context is different. We are already near full employment. So, this is not just about creating more jobs. It is about ensuring our people can take on the jobs that are created.
Because growth will come and new roles will emerge. But if our workers are not ready, if skills are not kept in pace, we risk a different kind of jobless growth – not a lack of jobs, but a mismatch between jobs and the skills our workforce has. Avoiding this means focusing on capability, not just capacity.
Businesses must transform jobs, not eliminate them. Workers must continue learning. The Government must support hope with strong systems and pathways.
When we do this well, growth will not leave people behind. It will lift them. My Deputy Speaker, in Malay, please.
( In Malay ) : [ Please refer to Vernacular Speech .] Mr Speaker, the discussion about AI today is no longer something distant from our lives. In Singapore and around the world, one thing is clear – AI brings hope, but also concern.
Within our Malay/Muslim community, these concerns are real. Young people worry about the future of employment and whether opportunities still exist for them. Those in mid-career are anxious about whether their experience and skills remain relevant. And blue-collar workers – including drivers, technicians and general workers – wonder whether their jobs will be replaced by automation.
We must acknowledge these concerns. But at the same time, we cannot view AI only as a threat. We must see it as an opportunity – an opportunity to progress together.
Allow me to touch on three important points in the AI transformation.
First, on building skills to be part of this transformative force. AI will only become a force multiplier if we know how to use it. That is why we must be prepared to continuously upgrade ourselves – to reskill and upskill.
This is where it is important for us to make use of the support available – whether through Workforce Singapore (WSG) programmes, NTUC, or community initiatives, such as M 3 +. At M 3 + Pasir Ris–Changi, for example, we run various programmes to help the community enhance their skills and employment opportunities. These include Women-at-Work to help women return to the workforce, as well as the Career Marketplace held in Pasir Ris to open access to career opportunities and employment networks.
And these efforts do not stop there. Through the HashTech programme in Pasir Ris–Changi, we are beginning to introduce our children to AI, robotics and autonomous systems – including through activities such as Robot Wars. This is not merely an activity. It is an effort to build confidence, exposure and future-ready skills from a young age.
Regarding skills, the fear of AI should not prevent us from learning how to use it wisely. The concern that AI could diminish any deep understanding of the Malay language and culture, as well as other skills, is understandable – if it is used without understanding the basics and its limitations.
Like any other tool, what matters is how we use it. Mastery of and appreciation for the Malay language remains key in understanding its beauty, values, and meaning.
However, if used appropriately, AI can also help preserve our heritage. AI can assist in digitalising old Jawi manuscripts, producing batik designs inspired by traditional Malay motifs and facilitating the learning and translation of the Malay language. Even classic Malay films can be preserved for future generations.
Technology should not erode our identity. If used wisely, AI can help preserve the language, strengthen culture and carry Malay heritage into the future with confidence.
Allow me also to share a pantun.
Golden bananas brought to sea,
One ripens atop a chest,
If AI is used very wisely,
Culture is inherited, in the heart it rests.
(In English) : Mr Deputy Speaker, Sir, at its heart, this is not just an economic transition. This is a test of our social compact, a compact that must now be renewed for a new era where businesses commit to not just profits but to people, where workers commit to not just jobs, but to lifelong growth, and where the Government continues to stand with both, ensuring that no one is left behind.
If we can do this, if we can move forward together with trust, purpose and shared responsibility, this transformation will not divide us. It will make us stronger because when workers, businesses and the Government move together, we do not just adapt to change. We shape it, we benefit from it and we ensure that every Singaporean moves forward together. Mr Deputy Speaker, I support the Motion. [ Applause. ]
Mr Deputy Speaker : Assoc Prof Terence Ho.
4.54 pm Assoc Prof Terence Ho (Nominated Member) : Mr Deputy Speaker, I rise in support of this timely Motion. I would like to first declare my interest as the executive director of the Institute for Adult Learning at the Singapore University of Social Sciences.
AI, as we know, is a transformative technology. All of us recognise that it will profoundly transform business models across the economy and with it, the content and nature of work.
As a small, open and technologically advanced nation, Singapore must strive to be at the forefront of new technologies, especially one as significant as AI. In addition to being critical for our economy, AI also has considerable potential to help address Singapore's demographic and societal challenges. But whether we embrace AI or fear it, there is no escaping the impact it will have on our companies and workforce.
I will make four points in my speech.
First, jobless growth is not an option for Singapore as good jobs are integral to inclusive growth and to a fair and vibrant society.
Singapore's social compact is based on self-reliance through work. This means providing for oneself and one's family through employment and income. It has been said before in this House many years ago that "a job is the best welfare, and full employment is the best protection for the workers of Singapore."
In this social contract, the Government's role is to nurture a pro-enterprise business environment conducive to investment and job creation. Any Singaporean who is willing to work hard will have enough to meet his or her housing, healthcare and retirement needs through CPF savings. In practice, significant Government support is given in the form of housing grants, CPF top-ups and healthcare subsidies to support home ownership, retirement adequacy and healthcare assurance.
Singapore's socioeconomic model has evolved over time. We now have more extensive risk pooling through social insurance, complementing individual savings in meeting healthcare and long-term care needs. There is more structural or permanent social support in the form of the Workfare Income Supplement and Silver Support. And the Government, recognising the greater risk of economic and job disruption, has introduced income relief for the involuntarily unemployed through SkillsFuture Jobseeker Support, which has been discussed by Members of this House.
Yet employment and income remain central to Singapore's socioeconomic model and social compact. Recent advances in AI pose a challenge to this model. Globally renowned AI pioneers and industry leaders have warned of the possibility of mass job displacements arising from AI. Predictions of a so-called "white collar bloodbath" or a "jobs apocalypse" have stoked public fear even as other commentators have asserted that these fears are overwrought.
Transformative technologies in the past have indeed eliminated certain jobs, but they have created new jobs as well so that we are not all without jobs or leading lives of unlimited leisure.
Generative AI has raised particular concern because it can take over jobs and tasks associated with human skill and creativity, including cognitive tasks such as coding and data analysis and creative tasks such as writing and design. These tasks require skills built through years of education and training and are consequently well remunerated. They underpin many of the good jobs that Singaporeans aspire to.
While we should not underestimate the disruption from AI, there is still time to adapt and respond. That is because the extent of job disintermediation or displacement depends on the speed of technology diffusion, which has historically followed an S-curve.
I filed a Parliamentary Question in March asking whether MOM has detected any signs of AI prompting a slowdown in the hiring of fresh graduates. Indeed, many Members of this House have pointed to this concern. The response I received then, which was reaffirmed yesterday as well, was that employment rates for young degree holders has remained broadly stable.
This may be because the adoption of AI is still in its early stages for many companies here. It takes time for more firms to move from viewing AI as a mere project or productivity enhancement tool to fundamentally re-organising work processes around it. As the pace of AI adoption picks up, however, the benefits to firm productivity and profits will grow, but so too will the impact on jobs.
A related concern that many in this House have also discussed is that in the AI age, profits will increasingly accrue to technology companies and those who own shares in these companies, and less to workers as skills become commoditised.
While we must consider the need for greater social support and new channels of redistribution to keep our society inclusive, we must also continue to equip citizens to provide for themselves through good jobs and incomes.
Some of us may remember what President Tharman explained in an interview in 2015 at the St Gallen Symposium. He was describing Singapore’s approach, and I quote: "It is about keeping alive a culture where I feel proud that I own my home and I earn my own success through my job. I feel proud that I’m raising my family. And keeping that culture going is what keeps a society vibrant."
If we look around the world at how economies have developed, it is clear that the key to success is the emergence of a strong middle class enabled by education and job creation. Many countries with large natural resource endowments have not done well, because their focus has been on resource extraction, benefiting the few, rather than education and skills development, which benefit the many. It is also evident that Singapore succeeded precisely because people are our only resource. Singapore's economic development has been a story of inclusive growth – and that is the path we must stay on.
This brings me to my second point, which is that we must support workers to develop deep domain knowledge, learning agility and career resilience. To equip workers for the AI age, training in AI tools is certainly important. Familiarity with AI and understanding of the strengths and limitations of different AI tools comes with frequent use, tinkering and experimentation. But that, as many in this House have also realised, is only part of the answer. After all, AI tools are supposed to become more intuitive and easier to use over time.
The real value that people bring to jobs lies in deep understanding of domains or subject matter, which means all of us still have to put in the hard yards to learn and avoid "cognitive offloading" by creating productive friction in the learning process, whether in schools or at the workplace. AI can provide the scaffolding and can help personalise learning, but it must not become a substitute for thinking.
The key skill that many have identified is learning how to learn, being adaptable and resilient to changes. This means getting out of our comfort zones, continuously challenging ourselves, getting used to different tasks and work environments, and working in diverse teams.
This is everybody's responsibility: to take ownership of our own learning and career development, leveraging public funding and resources where available.
Employers too have a responsibility. The Singapore Opportunity Index developed by MOM highlights how employers can create opportunities for their workers such as through recruitment practices, career development pathways and job design. By identifying employers with a good track record of supporting career growth, employers will hopefully be encouraged to invest in their workers and help them chart their careers.
The third point I would like to bring up is that we must not only upskill workers for jobs, but also upgrade jobs for workers. As I mentioned in my speech at MOM’s Committee of Supply debate, there will continue to be strong demand for workers in areas such as healthcare and skilled trades, driven by an ageing population and the relative resilience of these roles to AI displacement.
However, such jobs have difficulty attracting Singaporeans as they are perceived to be less prestigious or rewarding, or perhaps less comfortable compared with white-collar office jobs. But the risk is that Singapore will become increasingly dependent on foreigners to take on essential work even as Singaporeans struggle to find jobs that meet their aspirations.
Today, over 40% of the resident workforce have university degrees. Jobs and occupations that have traditionally been regarded as non-degree jobs therefore have to be redesigned to be suitable for a broad range of workers, including graduates.
Pay is only part of the issue. The jobs have to be redesigned to tap on workers' "head, heart and hand" skills, so that they are more engaging for workers and more resilient to AI disruption. This may involve taking on greater professional responsibility, increasing the cognitive, analytical and innovation content of jobs, and creating more opportunities for interpersonal engagement in the jobs. AI tools can in fact support the upgrading of vocational and semi-skilled jobs by augmenting human expertise. By making the whole range of these jobs attractive to Singaporeans, we can avoid structural overqualification or underemployment. Enterprises can do so by centring job design on skills rather than credentials.
My fourth point is that Singapore should build up expertise as a global reference point for skills-first and human-centric job redesign. A skills-first approach means that workers are not pigeonholed by formal qualifications into certain jobs or occupations. Employers recognise that workers, regardless of their starting point, can upskill to meet job requirements. Likewise, the scope of jobs and occupations can expand to make fuller use of workers' skills.
AI is already fractionalising jobs – breaking them down into tasks, some of which are assigned to AI and others to human workers. It is important for job redesign to be human centric so that human workers can still make a valuable contribution, augmented by AI and technology, rather than have processes entirely automated with minimal human involvement.
With agentic AI now able to execute processes that span multiple job roles, it is no longer enough to redesign tasks within a particular job. Instead, organisations must look at redesigning end-to-end work processes. This is a capability that must be embedded within organisations as AI continually reshapes the nature of work. Singapore has the opportunity to set the pace in skills-first employment practices and human-centric job redesign, which will benefit both our firms and workers and is critical for our social compact.
The Institute for Adult Learning’s Centre for Skills-First Practices recently launched a series of skills-first working papers and accompanying roundtable discussions. They attracted much international attention, with participation from experts, policy-makers and industry practitioners around the world.
The impact of AI on work and learning is an issue that all countries and societies are grappling with, and no one has all the answers. It is a fertile area of research as firms and societies seek to adapt and transform. We are in a complex operating environment with no tried-and-tested playbook to fall back on. As one corporate leader I spoke with recently put it: we are learning and adapting mid-flight.
Both learning and job redesign in the age of AI must be iterative and experimental. We need the best minds in Singapore and around the world focused on this. Just as Singapore has world-leading research centres in new technologies like quantum computing, we should build up expertise and experience in adult learning and human-centric job redesign.
Today, there are nodes of excellence across the institutes of higher learning. For instance, the polytechnics and ITE are among the national centres of excellence for workplace learning. The Singapore Management University has set up a Resilient Workforces Institute, while the Singapore Institute of Technology has launched a Skills Assessment and Validation initiative. At the Singapore University of Social Sciences, the Institute for Adult Learning has an Adult Learning Collaboratory and a Centre for Skills-First Practices. With end-to-end expertise from research to translation, the Institute for Adult Learning can serve as a national focal point for adult learning and employability.
Together, we can build up Singapore as a thought leader and living laboratory in the areas of adult learning, career health, human-AI complementarity and skills-first practices. As a global innovator plugged into international networks of cutting-edge research and practice, Singapore can help to shape the future of work in a way that supports both economic growth as well as human flourishing.
Mr Deputy Speaker, the Motion before us powerfully expresses the commitment of Singapore’s tripartite partners to inclusive development that benefits both workers and firms. This is important because it can no longer be taken for granted that economic growth in the future will be accompanied by full employment and rising incomes. Beyond adapting to AI, our end goal must be to enable every worker to grow, to contribute and to find meaning in work.
The formation of the National AI Council and the Tripartite Jobs Council underscores this commitment. It is a whole-of-nation effort where everyone has a part to play.
Recently, an American journalist approached me to find out how Singapore is addressing AI and its impact on jobs. She noted that many major US corporations are under pressure to slash headcounts to boost profits. Indeed, there are expectations from shareholders and investors that management will replace workers with AI. When I shared with her that Singapore’s priority is equipping workers to work with and alongside AI, she described the contrast, in her words, as "stark" and "inspiring".
With our unique tripartite collaboration and the commitment of all partners, I believe that companies and workers can approach the coming AI transition with resolve and confidence.
Mr Deputy Speaker : Mr Alex Yeo.
5.09 pm Mr Alex Yeo (Potong Pasir) : Mr Deputy Speaker, AI can be scary. I recall the first time at work when I received a draft legal submission from my younger colleague who had some content generated by AI, the alarm bells went off in my mind. "Is the content reliable? Am I going to be taken to task by the Court for submitting this?" I was anxious and worried even though the content had been verified and put together with considered "human"-generated legal analysis.
This incident made me recollect what happened when I first joined the legal profession. I would find written memos on my desk with instructions on matters from a particular Senior Partner. When I asked once, whether an email might be easier, I was told that an email was not reliable and that with a written memo, he was assured that I would receive the message. We all know better than to argue with our bosses but eventually, the written memos did move onto become emails, perhaps with the realisation that the email instructions would reliably reach me, even after office hours.
As with each industrial and technological transformation in history, be it the steam engine, electricity, digitalisation with personal computers, connectivity with the Internet and now AI, change and transition always bring with them anxiety and the fear of the unknown. Humans are wired to be untrusting of that which we cannot control – and maybe even rightly so.
With AI, we feel this acutely in Singapore. Many Members in the House have spoken about these concerns and the anxieties felt by Singaporeans from all walks of life. As the adage goes, "change is the only constant". Fear of the unknown is a natural reaction, but we should use it as a galvanising force to re-think old ways, learn new ones and as a result, seize new opportunities and grow.
This Motion is therefore a timely one. It recognises that harnessing AI to grow, is a double-edged sword. On the one hand, while AI can be a driver for the next phase of Singapore’s economic development; on the other, if its development and deployment is left unfettered, it can lead to societal ills such as job displacement and widening inequality.
As the Prime Minister pointed out during his Budget Statement, AI is but a "tool". How we harness it and manage its deployment will shape our economy, our jobs, our lives. Our approach to AI-enabled growth, as the Motion states, therefore, must be anchored in fairness, resilience and opportunities for all. The Motion resonates because it is about putting people at the centre of Singapore's approach to AI-enabled growth. Growth must be inclusive and benefit our people. We cannot have jobless growth at all costs.
Sir, awareness of AI's disruptive nature to the workplace and to our workforce is extremely high. The anxieties and concerns of our workers and PMEs across industries are not hypothetical. They are real, even quantifiable. A recent NTUC survey found that more than half of our PMEs feel an urgent need to upskill just to stay relevant. Nearly a third are actively anxious about being replaced. Other studies have shown that half of Singaporeans fear that their roles could be automated, and many are concerned that AI will benefit the bottom-lines of corporations more than it benefits the everyday worker.
Our young graduates entering the workforce for the first time face a daunting reality. As AI automates routine execution, employers are raising the bar, demanding higher-order critical thinking and AI collaboration skills from the outset. Left on its own, the AI disruption could very well lead to unfair outcomes. Organisations and individuals who manage a head-start or transition well, will benefit exponentially, while those who do not, will be left behind.
For years, we have proactively anticipated a disruption like this, by investing and imbuing in Singaporeans the value of lifelong learning and the need to regularly upskill through programmes, such as SkillsFuture. This has put us in a good starting position as we push hard on AI adoption, like the National AI Impact Programme that aims to support 10,000 enterprises and help 100,000 workers become more AI fluent.
However, AI fluency and literacy among the general population is also equally important.
On this I would like to raise two points.
First, you will recollect my personal anecdote at the start of my speech. Fluency with any tool, be it computers or smart phones in the past or AI today, is about building confidence.
Mr Lim Boon Heng, a former NTUC Secretary-General shared with me, his experience with the Government's computerisation efforts in the 1980s. It was a strategic decision at the time, but workers were afraid of computers. So, the Government designed computer appreciation courses that were rolled out by the NTUC, using early Apple computers. Familiarisation with the use of the keyboard and for those who remember, playing games like PacMan. Workers slowly got past the fear and embraced the familiarity. Importantly, the key message must be, this is a tool that can help you do a job, better, faster. Now, that new tool is AI.
While it is vital that we upskill our workers to leverage on the AI tools that are relevant for their respective workplaces, AI fluency and literacy should be a national endeavour that is available in our schools, our Community Clubs and even our Active Ageing Centres.
The idea is to embed AI fluency and literacy as a part of life, be it in creating a simple e-greeting card with an AI tool that creates moving graphics or building a billion-dollar company using bots. Only then can we envision an entire people advancing collectively together in an economy with AI-enabled growth.
If embracing AI is a national strategic move, then we should roll out a national AI literacy and fluency programme for all Singaporeans.
Second, supporting workers to become AI fluent should be more than just providing them with the AI tools and the know-how on using them. We must also equip and afford workers the time and ability to learn how to apply the AI tools in their workplaces.
Getting young graduates past the door, and securing older workers and PMEs in new roles are imperative, but if AI is to transform our economy, we must ensure that our workforce learns effectively through deeper learning while persuading our businesses and organisations to create workplaces and systems that allow our workers to test their new skills, learn from mistakes and improve. Learning must be coupled with building capability.
At its broadest and most pervasive use as a tool, AI can generate content, summarise and answer in seconds. We need a workforce that not only can use the AI tool to obtain these outcomes, but to work with AI as a collaborative partner while applying judgment, reason, creativity and context to drive high-impact value – the human elements.
Prof Er Meng Hwa, in a recent Business Times article advocating for deeper learning, gave the example of Micron Singapore, whose in-house AI upskilling initiative did not stop at awareness but allowed employees to use AI tools to extract insights faster, analyse risks better, automate route tasks, plan projects more effectively and improve decision-making.
How do we persuade more businesses and organisations to devote the time and precious resources to obtain these objectives together with us? Upskilling our workforce to ensure that they have the necessary credentials and knowledge may get them past the door, but it is no guarantee that businesses and organisations will employ and train workers in the way that will build on and optimise their AI capabilities.
This Government has been deliberate in focusing our policies on long-term advancement rather than on short-term gains. I am therefore confident that the AI National Council, led by the Prime Minister, will lay out plans and initiatives on the same basis.
In fact, we already have, within our system, the ability for the Government, unions and employers to collaborate closely to achieve the long-term objective of ensuring that our workforce is effectively equipped with the knowledge, skills, deep understanding and practical application outcomes.
Our tripartite model, where all three partners, having built trust, mutual respect and equal partnership over decades, working hand in hand through this new phase of AI adoption, will help us to ensure that our workers, our businesses and our economy can seize new opportunities and advance together.
Before I conclude, I wish to support the proposal by NTUC Secretary-General, Mr Ng Chee Meng, for the setting up of a market intelligence and foresight system contextualised to the Singapore market. We can appreciate how useful it would be to draw insights from information, data points, analysis and the research of tripartite partners to sense-make for early signals, coordinate responses and provide proactive early intervention, where necessary.
That said, while supporting our workers to retrain and move on to different roles to prevent displacement is important, we should also explore the possibility of using information from the same system to identify how the Government can incentivise AI startups to create new jobs and opportunities for our workforce. For example, it has been reported that AI movie production startups in China are supported with incentives and financial support. They are reported to have pioneered micro-dramas or vertical dramas that is a new entertainment format making waves globally.
Sir, if we are to leverage on AI as the next phase of Singapore’s economic development, we must transform, not only our economy but also the lives of every worker, every Singaporean. Mr Deputy Speaker, I support the Motion.
Mr Deputy Speaker : Mr Patrick Tay.
5.20 pm Mr Patrick Tay Teck Guan (Pioneer) : Mr Deputy Speaker, Sir, I rise in support of the Motion. I would like to thank my NTUC sisters and brothers in this House for coming to support this Motion.
Knowledge workers like PMEs are highly exposed to AI, unlike earlier waves of automation that mostly impacted the rank and file. Many of our middle-income PMEs belong to a sandwiched or under-served group. They are expected to perform like the top but are less protected than the bottom and they face increased competition from foreign PMEs. Mid-career PMEs are sandwiched between younger and older dependents, and cannot afford to lose their jobs. Yet they can take a longer time to find new jobs when displaced due to their higher income and age, often at the expense of a pay cut.
We perceive PMEs as privileged and adaptable, who have resources and can take individual responsibility to upskill and also to cope with setbacks. This assumption no longer holds true. AI is set to both augment and disrupt job tasks across all PME sectors at all levels. We must recognise AI as a transformative technology that has the potential to create new opportunities and shared prosperity but also the potential to widen inequality by concentrating wealth and power at the top, especially if guardrails are disregarded for the broad middle in the race to adopt AI.
To this end, I submit that Singapore's approach to AI must be human-centred while AI-enabled growth must be worker-centred. This means that we commit to what I call the "3 Es”. The "3 Es" are equitable growth, enhanced protections and engaged workforce.
First "E", equitable growth. When we talk about AI-driven jobless growth, this does not necessarily mean mass unemployment. However, it may mean that the gains from AI may not trickle down to the broad middle. Earlier this year, the Economic Development Board (EDB) announced that the expected number of jobs to be created fell to 15,700, the lowest in at least 20 years, despite having attracted more investments than the year before. This indicates that while we may not see jobless growth, we may well see growth with less jobs. Jobseekers could be competing for fewer vacancies, underemployment could rise and wages could stagnate. Young graduates may find more challenges securing full-time employment compared to before.
Globally, we have already seen a wave of companies, especially those in the tech sector, announce wide-ranging job cuts citing AI as a cause.
Equitable growth means that gains from AI-enabled growth will be shared with workers, in the form of better wages, welfare and work prospects. As a start, we need to raise our standards for what constitutes fair and responsible retrenchment, such as by requiring early retrenchment notifications, supporting unions to negotiate for retrenchment benefits according to industry norms and designing AI grant incentives with conditions requiring employers to demonstrate efforts to meaningfully re-deploy workers whose roles and tasks are taken over by AI.
As Singapore aims to attract the world's top AI talent to our shores, we must also ensure this builds up and strengthens our Singaporean Core. I ask that the Government encourage reciprocity through programmes, like the Capability Transfer Programme, so that knowledge and expertise flow to our local PMEs. At the same time, the Government can also invest in homegrown AI talent by sending them on training stints to top AI companies overseas through programmes like the Overseas Market Immersion Programme. Bringing in global expertise and developing our own global talent pipeline are two sides of the same coin. Both deepen the capabilities of our local workforce.
Second "E", enhanced protections. Inevitably, some workers will be impacted in the transition to AI. We cannot leave them to sink or survive on their own. They will need career guidance, financial support and grace to bounce back stronger. I thank the Government for launching the Jobseeker Support Scheme for the involuntarily unemployed workers to benefit from transitional support for up to six months while they train or search for their next job.
The income threshold for Jobseeker Support Scheme is currently set at $5,000 excluding CPF, which would only cover less than 20% of resident PMEs. I hope that the Government will consider raising this threshold to the gross median resident PME income, currently at $8,400 as of 2025, or consider some other suitable schemes of equivalence to address the needs of impacted PMEs. By the same token, the Government can also consider allowing those with HDB housing loans a temporary deferment for up to six months, especially if they are unable to meet the mortgage payments to ease immediate cashflow needs.
Singapore has taken a considered, framework-based approach to AI governance. We are not behind. We are being deliberate. Through IMDA's Model AI Governance Framework, the Cyber Security Agency’s addendum to secure agentic AI systems and voluntary testing toolkits, we have encouraged innovation and responsible adoption.
But as AI moves from assisting decisions to making them, including in hiring, promotion and restructuring, we must keep pace. Other advanced economies have already taken the lead to explicitly classify employment-related AI as "high risk".
This matters greatly to workers. Will recruitment AI automatically rank candidates with disabilities lower? If HR relies on an AI tool to shortlist who stays and who goes during restructuring, what recourse does a worker have if the outcome feels unfair? Who is accountable in this case – the HR team, the employer who procured the AI tool or the developer who built it?
These are not hypothetical concerns. A joint Singapore University of Technology and Design (SUTD)-GovTech study found that large learning models can reliably guess personal characteristics, like gender of candidate from other data points like hobbies and volunteering, even when resumes are anonymised.
We are fortunate to have passed the Workplace Fairness Act. Its principles – that employment decisions must be fair and merit-based and that every worker deserves fair access to good jobs – should apply regardless of whether those decisions are made by a human or by an algorithm.
But while these principles are technology-agnostic, our current anti-discrimination levers have yet to explicitly address how they interact with AI-mediated decisions. As AI adoption accelerates, employers need clarity on what responsible use looks like, and workers need assurance that existing protections travel with them into an AI-enabled workplace. We should close this gap, not with regulations that stifles innovation but with clear, practical tripartite guidelines to ensure the just and fair transition. Singapore's strength is our tripartite approach and we should leverage it.
To that end, I ask that the Government consider the following. First, that employers who adopt employment-related AI be guided to conduct risk assessments proportionate to the level of impact and ensure meaningful human oversight. Second, that HR professionals be supported with training and self-assessment tools to use AI responsibly and identify biasness. Officers involved in data governance and cybersecurity can also update their skills and knowledge in this evolving area as more companies adopt AI that is integrated with enterprise data. And third, that workers be given transparency. They should know where AI is being used in decisions that affect them, what guardrails are in place and how existing avenues of redress apply. And fourth, that we explore going upstream, working with AI vendors and developers to ensure that the underlying software meets some baseline principles of fairness and transparency before it reaches our workplaces. The International Labour Organization has started this journey and Singapore, with our tripartite DNA, can be a frontrunner in this space.
Third and final "E", engaged workforce. Workers master AI and figure out how to embed AI into their workflows, not the other way round. Ultimately, an organisation cannot be run by no workers and AI alone. Simply imposing AI from the top will not produce results and can even create resistance and sunk cost. Workers are end-users and experts in their own workflows. They know where AI adds value and where it falls short. If you want AI to work, you have to ask the people who do the work.
We already have a proven mechanism for doing exactly this – our NTUC's CTCs. CTCs bring management, unions and workers together at the company level to co-design transformation plans, pairing technology adoption with job redesign, skills upgrading and better wages.
Let me share one example which the Prime Minister shared during the May Day rally. At Tan Tock Seng Hospital (TTSH), where our Healthcare Services Employees' Union (HSEU) worked with management through the CTC to roll out a Smart Scheduler that could handle multiple shift patterns and cut rostering time from more than 90 minutes to under 15, so that nurse managers could spend more time on other core work instead.
Senior Enrolled Nurse Lilian Teng, 69 years old, who has worked at TTSH for the past 19 years, put it simply: with technology making work less physically demanding, she can continue working effectively for as long as she remains healthy. That is what an engaged workforce looks like.
But company-level efforts alone will not be enough. With the formation of the Tripartite Jobs Council, the Government, employers and unions can now coordinate sectoral transformation with workers at the centre, ensuring that AI training, job redesign and transition support are shaped by those on the ground, not just decided from the top.
Critically, the Tripartite Jobs Council can level up the CTC ecosystem to extend its reach beyond large employers to SMEs that may not have the resources to navigate their transformation alone. CTCs engage workers at the company level. The Tripartite Jobs Council will do it at the national level. Together, they ensure that AI transition is not something done to our workers, but with them. In Chinese, Sir.
( In Mandarin ) : [ Please refer to Vernacular Speech .] Mr Deputy Speaker, AI is here – and it has arrived fast and with force. But we must ask: when AI generates wealth, whose pockets does it go into? I raise three points.
First, share the cake fairly. AI creates wealth, but that wealth cannot flow only to employers. Workers help bake the cake and they deserve a slice of it too.
Second, if jobs are lost, we cannot leave people to fend for themselves. The eligibility criteria for Jobseeker Support assistance can be broadened, so that more PMEs can benefit and have greater security. We must put up the umbrella before the rain comes – prepare for the storm before it arrives.
Third, walk this road together. The AI transition cannot be decided by employers alone. As the saying goes, three cobblers together are better than a Zhuge Liang. The tripartite partners, unions and employers must work together, only then can we go far, steady and fast.
Share the cake fairly and walk forward together. That is the AI future that belongs to every Singaporean – one that puts people at its heart.
( In English ): To conclude, there has been much said about AI as a double-edged sword. This metaphor has been used ad nauseam, but this is not untrue.
AI's impact on the broad middle means that it is a "once-in-a-generation technology" but could also be a once-in-a-generation divider that concentrates gains to those who control AI while displacing the very same workers who helped to design, implement and build it.
Equitable growth, enhanced protections and engaged workforce. These are three principles that must guide us in the way forward if we hope to anchor AI as an enabler in our AI transition in fairness, resilience and opportunities for all, because every worker matters. Mr Deputy Speaker, Sir, I support the Motion.
Mr Deputy Speaker : Minister of State Jasmin Lau.
5.34 pm The Minister of State for Digital Development and Information (Ms Jasmin Lau) : Mr Deputy Speaker, I have listened carefully to Members today. There is genuine concern across the House about what AI will mean for our workers. These concerns are real. The Government shares them.
We cannot slow down the development of AI. But we will not leave its outcomes to chance. We will work hard to secure a different deal between the companies that prosper here and the workers whose effort makes that prosperity possible.
Where companies benefit from operating and growing in Singapore, we will expect a fair deal for workers. Not just in words, but in how jobs are designed, how people are trained and how gains are shared. Where public resources and policies are used in support of business transformation, we expect companies to deliver clear and meaningful outcomes for workers.
In my conversations and engagements across the Ministry of Digital Development and Information (MDDI) and MOE work and through the Economic Strategy Review, the same concerns come up again and again. Will my job exist in five years' time? Will AI widen inequality and leave the vulnerable behind? If AI makes companies more productive, will workers share in the gains?
These are not unreasonable fears. They come from people who have worked hard, built up skills and experience over time, and now sense that the ground is shifting beneath them. I will take each of these questions in turn.
First, on whether today's jobs will continue to exist. Let me be honest. Some roles will change substantially. Roles built primarily around repeating the same steps are the most exposed. This is not a verdict on the value of the people who do that work. It is a signal to us in government and to employers that we need to act now and not after the disruption arrives. And act, we will.
But AI is more than just a technological advancement that replaces jobs. At the same time, it is opening up entirely new ways of working and new kinds of roles that did not exist before.
Some academics have described AI as an "invention of a method of invention". It expands the space of problems that can be solved, the products that can be built and industries that we can create. A small biotech team in Singapore can run experiments that would have required a national lab decades ago. A solo founder can ship software that took a hundred-man firm to deliver three years ago.
So, competition sharpens, but the frontier also moves outward. That is why the Economic Strategy Review Committee that Senior Parliamentary Secretary Goh Hanyan and I co-chaired, focused on identifying new areas where Singapore can use AI to build a real competitive edge. And the Prime Minister's National AI Council will take this forward.
Members have pointed out the impact of AI on PMEs, as AI automates routine and analytical tasks. The Economic Strategy Review team recognised this, which is why helping businesses and workers proactively navigate the transition was the focus of the committee chaired by Minister of State Goh Pei Ming and Minister of State Desmond Choo.
For displaced workers, the committee studied how the Government, employers and unions could offer more timely help. As mentioned in our mid-term update, the Economic Strategy Review is studying ways to encourage earlier retrenchment notifications as raised by Mr Ng Chee Meng.
On PMEs specifically, the committee recognises that they may face greater job uncertainties and will recommend more targeted support. This includes considering enhancements to the Jobseeker Support Scheme, as Mr Patrick Tay suggested, and tapping on private sector expertise to strengthen placement support for this group.
For workers at risk of displacement, the Economic Strategy Review will recommend practical ways to help them move into more resilient roles with stronger demand. We will identify sectors with sustained labour demand and lower AI displacement risk, and we will work with unions and employers in those sectors to create clear, supported entry points for workers making the transition. We must make these pathways walkable and not just visible.
To illustrate, a mid-career worker in a routine administrative role, for example in data entry or customer service, could be worried about AI displacing him. With job facilitation and reskilling support, the worker should be able to move into a sector where there are roles that build on his existing skillsets. For example, the worker could explore adjacent roles in healthcare administration. This is where we are seeing robust demand given the growth of our population healthcare needs, and healthcare requires uniquely human skills that are more resilient to disruption.
All this requires more than courses. It requires employers, unions, training providers and placement support working closely together, so that workers do not fall through the cracks during transition.
No government in the world has all the answers to this transition and I would be wary of any that claims otherwise. What we in Singapore can commit to is this: we will not wait for perfect solutions before acting. We are starting now and we will adjust our efforts along the way.
Second, on inequality. Members are right to worry. Technologies that amplify capability can also amplify gaps between those who adapt quickly and those who struggle to keep up. As Mr Mark Lee points out, some risks we face are that productivity gains accrue more to those already ahead, while the bottom of the career ladder may face erosion.
Our response is to raise the floor and widen the door. This means starting earlier: building AI literacy into our schools, so that all students develop confidence with AI and not just those who have access to resources.
Currently, every ITE and polytechnic student is already taught AI literacy as part of their course, and we are now bringing AI literacy and safe AI tools into primary and secondary school classrooms. This means that all students, regardless of economic background can learn about AI safely. They can also learn how AI can benefit their learning, such as to help them refine their ideas, and they also learn when they should not use AI.
As Minister Desmond also pointed out earlier today, we are committed to supporting students who may not have strong family or parental supervision and support. While AI literacy in school will give them a good and strong foundation, we must continue to develop partnerships with the community and the self-help groups to make sure the supervision and support continue outside of school.
Learning must continue beyond graduation. From the second half of 2026, all of our IHLs will offer selected AI-related courses at significant discounts for their alumni, for a period of one year.
For workers already in the workforce, Singaporeans who complete selected AI training courses will receive six months of complimentary access to premium AI tools. We will track take-up and usage to see if we need to do more.
Every Singaporean, regardless of starting point, should have the chance to experiment with AI tools and to build fluency.
The third question is the hardest and the most important. Will workers share in the gains? We should be clear: this does not happen automatically. Left on its own, technology can lead to very uneven outcomes. That is why this is not just a market question. It is a question of how we shape norms and expectations in our economy. So, let me set out clearly what we expect.
Companies that benefit from AI should invest in their people, not just in technology. That means training as many existing workers as possible and not just hiring new ones. It means facilitating their employees' access to frontier AI tools, creating communities of practice and incentivising learning and upskilling.
It also means redesigning jobs in close consultation with workers, as suggested by Ms Yeo Wan Ling, so that people can work alongside AI, using judgement, context and experience, rather than treating workers simply as a cost to be reduced. And where roles do change or disappear, it means making a serious effort to redeploy and reskill workers within the organisations before turning to retrenchment.
We are not just asking our companies to do national service. We are asking them to do what is in their own long-term interest. In an AI age, human instinct and intuition will remain key. We all know that when we work with AI, we need to steer it. Ask the right questions and apply judgement as we refine the output iteratively.
It is not one shot. If you do not develop people who understand the context of your organisation and use this knowledge to reinforce your AI systems, you will be left with a very shallow and hollow company in future. If companies here replace humans completely with AI, they will find themselves, in future, to have no competitive edge, when AI is available to all companies. They will also find themselves at the mercies of AI companies. So, what we are working towards, is an approach that best positions our companies for sustainable growth in the long term.
Mr Saktiandi Supaat brought up the need for balanced regulatory approaches that do not disincentivise AI adoption. Indeed, we will not seek to legislate our way to good outcomes. That has never been Singapore's primary approach. But we are equally clear that "voluntary" cannot mean "optional in practice".
Where public resources are deployed, we will ask for worker outcomes. We will work with companies to meet these expectations. Where there are persistent gaps, we will review how our support is applied. We will discuss with our tripartite partners on how this can be done fairly and effectively, in a way that incentivises companies to invest in training, job redesign, redeployment and placement.
If we do this well, we will be able to create and sustain good jobs in the AI age. A good job is not just a job that exists. It is one that allows a worker to progress. It should pay fairly and reflect the productivity gains that technology brings. It should build skills that remain relevant, including as part of routine on-the-job training, so that workers are not stuck doing tasks that are easily replaced by automation. And it should give workers a sense of dignity and agency, not reduce their role to simply following instructions generated by machines.
We have seen that when there is strong commitment, this is possible. At PSA, AI and automation have helped deliver record cargo volumes. At the same time, the company reskilled and redeployed more than 2,000 workers into higher-skilled roles. And they continue to hire thousands more, because they are growing faster than the competition.
To Mr Andre Low, I would say: automation and augmentation are not mutually exclusive. Protecting a worker can mean being intentional about automating tasks that are repetitive and physically demanding and upgrading the skills of the same worker so that technology can augment his capabilities as he takes on a higher value role.
Even smaller businesses are playing their part. Take for instance our local pawnbroker Maxi-Cash. In the past, a customer wishing to trade in jewellery would interface with a sales advisor, who would pass on their case to a valuer to assess the authenticity of the jewellery. Maxi-Cash enhanced this process by reskilling 25 of their sales advisors to use an AI-enabled authentication system, which can accurately assess the composition of jewellery in just five seconds. Now, these sales advisors can complement the existing pool of valuers, relieving their workload and reducing the customer wait times. This is the kind of responsible transformation we want to see in Singapore as the norm, not as the exception.
Mr Deputy Speaker, I have listened carefully to the many suggestions and perspectives shared in this House today, from Members on both sides. We may differ on specific policy ideas or on how particular measures should be designed, funded or sequenced. That is the nature of democratic debate, and it is very healthy.
But I believe there is broad agreement across this House on a fundamental principle: that the gains from growth and progress must be shared fairly and broadly with all Singaporeans. This should not be a matter of party or ideology. It is a principle that we must uphold together as Singaporeans.
So, let me say this plainly. If Singapore succeeds with our AI ambitions – and we should never assume success is automatic, because it will require sustained effort, difficult choices, adaptation and perhaps, some good fortune too. But if we succeed, then the Government will ensure that the benefits are widely shared.
The gains must not accrue only to those who already have capital, advantages or access. They must translate into better wages, better opportunities and greater security for all Singaporeans. The best protection for workers is not only redistribution after disruption. It is shaping how gains are created and shared from the outset and ensuring that Singaporean workers retain agency within an AI economy.
This Government has been able to deliver these outcomes over decades of Singapore's development. And we are determined to continue doing so, as we navigate this AI transition. Our policies have never been static. We have continuously adapted, refreshed and strengthened them as circumstances changed. And that discipline will continue.
Ultimately, every Singaporean should be able to look at what Singapore has built and say, "I have a stake in this progress. I have a share in this growth. This future, belongs to me and my family too."
This shared commitment is also what makes Singapore's approach to this transition distinctive. Our strength is not just technology. It is the way we work together, across Government, businesses and unions.
To workers watching this debate, I want to say this directly to you: the Government is on your side, and we are acting before the disruption reaches you, not after. You will not be doing this alone. Our commitment made in this House today is our commitment to you.
To our business leaders: AI gives you powerful new capabilities. But how you use them will define your company's future, and your relationship with the people who built it alongside you. The companies that will lead in 10 years are not those that stripped costs the fastest, but those that built stronger teams by combining human judgement with machine capability.
But I want to be clear about something else as well. Not every business needs to adopt AI, and not every pursuit needs to be seen through the lens of AI transformation. There is real value in things that are fully human created, and that value may grow, not shrink, as AI becomes more prevalent.
When everything around us is auto-generated, optimised and scaled, the things that are not will stand out. The live performance and encore that cannot be repeated. The hand-thrown ceramic bowl that carries the mark of a human hand. The meal prepared with care and craft, not just consistency. The conversation with the calligraphy master who has spent a lifetime honing his art.
I think we will see a revival of appreciation for these things. And Singapore should not just acknowledge this, we should embrace it. Our artisans, our performers, our craftsmen are not swimming against the tide of AI. In a world saturated with AI-generated content, they may find themselves exactly where the world is looking.
Beyond the near-term transition, there is a longer-term question we must answer. What do we need to do now with our education system, to prepare our students for the future world?
We must accept that AI will continue to get better at the tasks which machines do well. All the more, we need to focus on what makes us distinctly human. The curiosity that asks a question nobody has thought to ask. The creativity that connects ideas across domains in ways no training data predicts. The empathy that reads a room, earns trust and knows when the most efficient solution may not be the right solution.
We often call these soft skills. In an AI age, they will become the hard edge of competitive advantage for our people and for Singapore. That is why we will review our education system, to make sure we develop these qualities with the same rigour and intentionality we have always applied to academic excellence.
We must continue to build strong foundations and make sure our students do not become overly reliant on AI shortcuts. Our human brains are muscles that require exercise, and genuine mastery – the kind that holds up under pressure and that AI cannot simply replace – comes from hard work, from practice and from deep understanding. So, it was good to hear Ms Eileen Chong agree with this and we thank her for supporting our approach.
But rigour and exploration are not opposites. The student who has truly mastered something is precisely the one with the confidence to venture beyond it. He will ask harder questions, take on problems without the obvious answers and he will develop interests that are genuinely his own. What we are building towards is an education system that demands both the discipline to go deep and the freedom to go wide. Not just because our students deserve both, but because Singapore's future depends on both.
This will not mean abandoning our standards. It will mean expanding what we count as excellence. A student who asks the unexpected questions, who pursues something deeply out of genuine interest, who can hold two contradictory ideas and work through them – that student is not behind. In the world that we are building, that student may be ahead of all of us.
We are committed to doing this, together with our educators, our parents and young Singaporeans themselves. Because if we get this right, if we develop a generation that is not just AI-literate but deeply human, then Singapore will not just survive this transition. We will be the kind of society that the next era of human progress is built around.
Mr Deputy Speaker, we have been honest today about what this transition will demand – of the Government, businesses and workers. Not every path will be smooth. Some will face real disruption and our responsibility is to ensure no one faces it alone. We will make AI work for Singaporeans. And we will ensure that as our economy grows, our workers move forward with it.
But I want to end where I believe our attention must ultimately rest – on the generation we are building. If we develop Singaporeans who are curious, creative and deeply human, people who can ask the questions that machines cannot and earn the trust that algorithms never will, then we will not merely manage this transition. We will define what comes after it. I support this Motion. [ Applause. ]
Mr Deputy Speaker : Senior Minister of State Desmond Tan.
6.00 pm The Senior Minister of State, Prime Minister's Office (Mr Desmond Tan) : Mr Deputy Speaker, Sir, I begin by declaring my interest as the Deputy Secretary-General of NTUC and the executive secretary of the Singapore Industrial and Services Employees' Union (SISEU), where I am closely involved in supporting our workers. Today, I want to continue to support our senior workers and reflect their concerns in the age of AI.
Let me start by sharing about Mdm Foo, a 55-year-old jobseeker who approached NTUC's e2i for assistance. After leaving her previous job of more than 20 years, she found that the job search process had changed quite dramatically. Even resume writing has changed. Resumes used to be written for people, but she found out that today, they are often screened by machines first. Job applications also moved to digital portals that are not so intuitive for people like her to navigate. So, she felt lost and uncertain.
Mdm Foo's experience is not uncommon among senior workers and reflects their anxieties about changing work processes. For some, AI presents true opportunities. For others, it creates uncertainty and anxiety. And for many of our senior workers, their experiences are shaped by three key gaps that I think we collectively must try to address.
The first is the access gap. While Singapore has made progress in closing the access gap for seniors, smartphone ownership among them still lags behind other age groups. In addition, seniors may have less access to AI tools.
I saw this first-hand at one of the AI workshops organised by my union SISEU for about 90 union leaders whose median age was about 53. While all of them used smartphones, many were trying AI tools for the first time. Understandably, there was some initial hesitation. But with simple, guided use cases, they quickly picked it up and were using AI to generate posters for the union's Family Day and membership campaign. Some even went to the extent to do banners for their families for birthdays and anniversaries.
As we ended the session, many leaders came forward to share with me that they enjoyed the session and now, they realise that it is not so difficult to learn to use AI tools. The only thing they asked for is that they wished the session could be longer and the font on the slides can be bigger for them.
This encouraged me because it shows that the issue is not a lack of willingness, but rather a lack of access and, to some extent, to some of them, it is about confidence, which we can overcome by curating and customising access to AI tools and making time for our seniors to learn and to increase their knowledge.
Second, skills gap. We see a gap in training participation, where MOM's 2025 report found that residents aged 50 to 64 had the lowest training participation rate at 44.5%, compared with about 60% for those who are below 40 years old.
It is easy to tell seniors to go and upskill, to go and take a course and reskill, but in reality, we know that with commitments, with bills to pay, with limited time and energy, taking the first step sometimes is not easy. Many of them, we must remember, have lived through repeated cycles of change and transformation, and may feel fatigued, uncertain or even question the relevance of more training.
These concerns are real. We must make training more accessible for them at a suitable pace, through practical and bite-sized modules, and make AI more relevant to their job skills.
The third is the opportunity gap. Even when senior workers are willing to learn, they may not have the same opportunities to benefit from AI in their actual jobs.
OECD's Employment Outlook 2025 highlights that across OECD countries, opportunities to learn by doing fall with age, where 62% of adults aged 25 to 29 reported such opportunities, but this falls to 45% among those who are 60 and above. In addition, a 2025 Pew Research Centre report found that 73% of workers who used AI at work were aged 18 to 49, while only 27% were aged 50 and above.
This is why we need to work with employers to give our seniors the opportunity to use AI tools and to reap productivity gains in their job. Senior workers bring valuable experience and with AI, these strengths can go even further. I have spoken about this no less than two times in this Chamber.
Research from the Stanford Digital Economy Lab supports this. It highlighted that senior employment in the US has remained resilient and may even have grown with the introduction of AI because they see that seniors bring tacit experience, knowledge and soft skills that enable them to increase productivity with AI.
To our senior workers, we understand your challenges and we are with you in this transition. Take the first step with us. Start a course, try a tool, learn from those around you and you can thrive in the AI economy.
Mr Deputy Speaker, the Motion before the House is important. It emphasises that growth must be anchored in fairness, resilience and opportunity for all and resolves to equip workers to seize these new opportunities. That must also apply to our senior workers.
NTUC recognises that long-lasting impact is best achieved through a partnership approach and has taken proactive steps through an AI-Ready SG initiative, which focuses on three key areas.
First, training and upskilling workers to address skills and access gaps. To close the skills gap, NTUC LearningHub has developed a comprehensive AI learning pathway, with three different levels for learners with different proficiencies: foundational training to build AI literacy and fluency, intermediate training tailored to sectors or job roles, and advanced training for those in deep tech and who want to develop deeper AI specialisation and capabilities.
I am glad to note that to date, there has been strong interest. Since February 2026, more than 4,000 workers have enrolled in LearningHub's AI courses. I am also very happy that 39% of them are seniors.
Dr Neo has suggested certification of competencies. This is something that the LearningHub has already been doing. For example, it works closely with companies to design AI courses aligned with the Government's skill frameworks and that are tailored to companies' needs. We also partner with industry leaders such as AWS and Microsoft to certify learners' competencies based on industry demands. We will continue to expand this to more sectors and more industries.
At the same time, under AI-Ready SG, we are also closing the access gap by providing union members with subsidies of up to 50% for AI premium subscription tools through NTUC's Union Training Assistance Programme (UTAP). I am happy to also let Members know that in the first of these AI tool subscriptions, NTUC's premium subsidy does cover a range of tools, including coding and agent-based tools such as Claude Code, Codex, Manus and others. I think there are a total of 20 or 21 tools.
We are offering this as part of membership privileges only because we are using the existing UTAP funding model catered for our members' training. But we will continue to review this, depending on the take-up and the interest over time.
We are also partnering sector agencies like the Civil Aviation Authority of Singapore to develop sector AI training pathways for our union leaders.
Second, we are supporting firms in business transformation and job redesign through NTUC's CTCs. To date, NTUC has formed 3,800 CTCs, embarked on over 900 business transformation projects, benefiting over 300,000 workers.
Let me share another example. I know you have heard many examples over the course of this debate. It is from the Evergreen Group, a local office and stationary supplier. Through a CTC Grant project with the Singapore Manual and Mercantile Workers' Union (SMMWU), it implemented an AI-powered e-ordering system to automate the ordering process and improve inventory management. With this new system, manual work such as order processing was reduced by about 60%. Workers could focus on higher-value tasks such as managing customer relationships and using data to optimise inventory.
As they became more productive, the company could handle 40% more orders and provide wage increments for its employees. This is what we mean by win-win outcomes – where businesses become more productive and our workers progress with them.
Third, we are improving job matching with new products and services to help workers access good jobs. Let me go back to Mdm Foo from the start of my speech. Through close support from her e2i career coach, she gained a deeper understanding of her skills and the new job market. Her coach also introduced her to NTUC's AI Career Coach and e2i's AI Interviewer. With support and encouragement, Mdm Foo could confidently use these AI tools to sharpen her resume and also practise her interviews before the actual session. I am glad to share that Mdm Foo has found a new role and has gained familiarity with AI in the process.
Mr Deputy Speaker, AI-Ready SG is one example of how we can realise the intent of this Motion. It supports workers to build AI skills, gives them access to tools to apply these skills and works with businesses to improve productivity and create new opportunities.
Looking ahead, we welcome companies and partners to come onboard as we scale our efforts and increase our reach to ultimately deliver better support for our businesses and for our workers.
Mr Deputy Speaker, these efforts are important and we are already seeing encouraging outcomes. But the scale of change brought about by AI is also significant, with great uncertainty and anxiety among workers. This is why tripartism, with its proven track record built over many decades of open communication and trust, is critical to this challenge.
Singapore has navigated major transitions before. In the 1980s, when computers first entered the workplace, workers were concerned that these machines could replace their roles in data entry and filing and companies were anxious about costs, skills shortage and disruptions to their daily operations.
But tripartite partners leaned forward. The Government invested in infrastructure and skills, including the National Computer Board, to drive nationwide adoption of IT. Employers stepped forward to transform their businesses with new technologies and redesigned workflows. The Labour Movement drove skills upgrading en-masse, organised workshops and seminars, preparing workers mentally and practically for change.
Because tripartite partners moved together in solidarity, firms became more productive, workers took on better jobs with better pay and Singapore strengthened its competitiveness.
Mr Deputy Speaker, the Tripartite Jobs Council will be a key platform to realise our shared aspirations for the AI era as laid out in the Motion. It will build on efforts across the Government, employers and the Labour Movement and enable partners to scale outreach, accelerate policy implementation and direct resources so that workers and enterprises can seize opportunities from AI.
The Tripartite Jobs Council will take on a practical and iterative approach. We may not have all the answers upfront, but we are clear that our deep trust built over decades of cooperation and shared goals will allow us to achieve our aspiration of inclusive economic progress. Mr Deputy Speaker, I will now speak in Mandarin, please.
( In Mandarin ) : [ Please refer to Vernacular Speech .] Mr Deputy Speaker, Singapore needs to harness AI to drive the next phase of economic growth. But more importantly, this growth must be built on a foundation of fairness, inclusivity and resilience for all.
Singapore's tripartite partners – unions, employers, and the Government – have worked closely together over many years, building deep trust that has allowed us to unite and overcome challenges together, no matter what difficulties we face. We will continue to support our workers in seizing opportunities and strengthening their competitiveness in the age of AI.
In February this year, the NTUC launched the AI-Ready SG initiative, actively encouraging workers to learn and master AI tools. This initiative helps them bridge the gaps in awareness, skills and access, so that they are better prepared for the AI-driven economy. As the saying goes, "Opportunity favours the prepared and success belongs to the most persistent." The age of AI is already upon us. I hope that everyone will join NTUC in actively upskilling, learning and applying AI.
( In English ): Mr Deputy Speaker, AI is the defining technology of our generation. But we face this challenge with strong foundations built over decades of tripartite cooperation.
To our tripartite partners, let us continue working closely together to help our firms transform and stay competitive, while supporting our workers to remain productive and to maximise opportunities.
To our workers, we will continue to support you to leverage AI.
And to our senior workers, your experience matters and it is my firm belief that it will be an advantage in the AI era. Take that step with us, upskill and learn, including from your younger colleagues and together, we can close the access gap, narrow the skills gap and expand opportunities for all. Because in Singapore, we have always believed that progress must be inclusive, that as we move forward, we move forward together, because every worker matters. Mr Deputy Speaker, I support the Motion. [ Applause. ]
Mr Deputy Speaker : Minister Tan See Leng.
6.16 pm The Minister for Manpower (Dr Tan See Leng) : Mr Deputy Speaker, Sir, let me begin by acknowledging what many Singaporeans are feeling right now – uncertainties, anxieties, a sense that the ground is shifting beneath their feet; the world feeling less predictable than what it used to be, with trade tensions, fragility in supply chains, wars in the Middle East and the sharp rise in oil prices.
Closer to home, Members of this House have spoken about something that weighs on the minds of all of our fellow Singaporeans: the anxiety that AI may erode our skills, our experience built up over years or even take over our jobs. This anxiety has been sharpened by news of large tech firms announcing lay-offs attributed to AI adoption.
These are legitimate concerns and we take it seriously. And a change of this magnitude is indeed unsettling. But AI can and will create opportunities that we cannot yet fully imagine. Of course, it will, at the same time, also bring about disruptions that we cannot fully foresee.
But there are, at the same time, early signs that gives us reason for cautious optimism. Recent global surveys show that two in three companies that made earlier AI-driven cuts are already rehiring. Why is that so? Because they found that AI could handle the predictable and the routine, but customers still wanted human judgement, empathy and the genuine connection that AI could not provide.
Let me offer a small illustration of my own. In the preparation for my speech today, my team used AI to help refine the work to me. It surfaced useful references, including our MOM's newly released study showing only about 6% of firms in Singapore have reduced headcount due to AI adoption.
But one thing it could not appreciate was the impact and the anxiety that many, many workers feel. It could not offer empathy, it could not empathise, it could not understand nuance, nor could it generate policy responses that capture the essence of what workers are really experiencing. And this is what no algorithm can replace.
[Mr Speaker in the Chair]
The Motion before us makes four commitments. The Government and MOM take each one seriously – all of them as foundations to build on and to go beyond.
Mr Ng Chee Meng spoke about the transformative impact of AI on our workforce. The Government has long recognised AI's potential. Our current efforts build on a strong foundation of work already done in this space. We formed our first National AI Strategy in 2019, well before the introduction of ChatGPT and we embarked on national AI projects in areas such as education, healthcare, logistics, security and municipal services.
When large language models exploded onto the scene in late 2022, making AI accessible and general-purpose, we refreshed our strategies with the National AI Strategy 2.0 in 2023 and developed plans to invest over $1 billion in AI compute capabilities, talent and industry development. And this included establishing AI Centres of Excellence and growing the number of AI practitioners.
As AI picked up speed and interacted with major shifts in our external environment, we convened the Economic Strategy Review last year to sharpen our response. And more recently at Budget this year, we formed the National AI Council chaired by our Prime Minister to drive the practical transformation of our economy using AI.
At every step, we have acted proactively with our tripartite partners to drive concrete action and transformation across the sectors. As a result, while we will walk into uncharted waters into an uncertain future, we can do so with some confidence and we are not totally unprepared.
Various Members have raised concerns over the impact of AI on job displacement and many have also put forward thoughtful suggestions on how we can better support workers and businesses through this transition. We hear your concerns and we welcome the suggestions raised by Members on both sides of the House.
There is, in fact, broad agreement across this House on what we are trying to achieve, which is inclusive growth for all in this AI transition. Where we may differ, is in how we get there. Our approach has always been to invest in our people, keep our workers economically valuable and shape how the gains from AI are created and shared. Rather than dwell on fears, on apprehension, we want to be able to inspire and motivate our workforce to continue to grow.
Mr Gerald Giam and Mr Andre Low highlighted structural threats to inclusive growth. I appreciate the seriousness with which they have engaged on this issue. We recognise the conundrums. The question is: where and how to intervene?
Mr Giam proposed a National AI Equity Fund to pay every Singapore Citizen $500 funded by the companies that benefit from AI. Mr Low similarly proposed a payout for those displaced through redundancy insurance. I recognise the need to strengthen our systems to ensure that no one falls through the gap in this transition. And I agree that the broad sharing of productivity gains does not happen automatically, because markets alone will not guarantee good social outcomes.
Let me be clear. The Government has always known this and has always been acting on it. Both Mr Giam and Mr Low's proposals rest on a more pessimistic premise, that Singaporeans are essentially passive passengers in the AI transition, without agency to seize the opportunities and can only rely on support for a journey they cannot steer.
I cannot hold on and I will not hold on to such a premise. Both your proposals are not empowerment. To me, it is a settlement. Resigned to the fact that mass displacement is inevitable and that the best we can do is soften the blow. We should have more confidence in the tenacity and the adaptability of our fellow Singaporeans.
Redistribution alone is insufficient if workers are excluded from the economy. Singapore's tradition has been to invest in people rather than to compensate them for the circumstances; and that is our true policy tradition, rather than what the Member, Mr Low, had described.
The better use of any surplus generated by AI adoption is to fund accessible and effective upskilling that amplifies Singaporeans' value. And to that end, the Government has spent over $10 billion over the last five years on local workforce initiatives.
The choice before us, Members of the House, is between two very different visions. One says you get a handout and then, with that, a small share of the pie that the machines produce. Whereas, on the other hand, we feel that you deserve to grow the pie with the machines and share in our economic prosperity through good jobs and good wages.
The first vision initially may seem generous, but ultimately, caps and diminishes your broader end objectives. The second demands more from the Government, more from employers and also more from workers, but it treats all of our fellow Singaporeans as capable adults, with futures worth investing in, not as a population to be managed through transfers.
And Members of the House, I believe the second vision is possible, because as AI transforms how we work, some jobs will evolve. Some jobs may disappear, but if we are able to get everyone on the same ship moving together, I believe we will prevail. And as Mr Mark Lee and Ms Yeo Wan Ling said, it also creates new opportunities for businesses and workers. Our duty and our focus will be to help all of our workers and our businesses to seize them.
So, we should never build the case in this debate on angst and on apprehension. Our approach is not to fear the future, but let us forge the future ourselves, because that is the true Singapore spirit.
We have seen through each wave of technology and economic restructuring with this very spirit, but we are not complacent. MOM is closely monitoring AI's impact on our workforce. Our inaugural survey of the firms shows that AI is currently augmenting, rather than replacing labour in Singapore. Only about three in 10 firms have adopted AI currently, and amongst the firms that have adopted AI, only a small minority, about 6%, reported reduced headcount.
More commonly, firms are redesigning jobs, they are creating new AI-related roles, indicating that AI is changing how work is done, how they are being re-organised, rather than reducing jobs. And seven in 10 firms using AI are already seeing productivity gains.
However, as I have said, we should never be complacent. We must be prepared that as AI adoption gains pace, momentum and scale, the impact on jobs would be greater. And that is why we constantly prepare ourselves.
We have the goal to enable more businesses to succeed. At the same time, where their workers use AI to do better jobs, rather than be replaced by AI. Where their work becomes more fulfilling, more meaningful, not less. And where the benefits of AI are shared between the businesses themselves and the workforce.
For workers seeing a more flexible pace of work, AI can enable new forms of flexible work and fractional work done by small teams or even "solopreneurs". Beyond flexibility, AI can also reshape who participates in our workforce, including seniors, as Ms Poh Li San spoke about. And we will explore how to scale flexible work models through the Tripartite Workgroup on Senior Employment.
Dr Hamid Razak spoke about the hesitation and anxieties he heard from the ground, especially among older PMEs who wonder if their skills still have a place. And Mr Yip Hon Weng also asked for businesses to be better supported. Let me share what Government is doing to prepare individuals and businesses for this transition.
First, we are reforming our workforce and skills support system to deliver more timely and effective support. As I shared after a five-hour debate yesterday on the Second Reading of the Skills and Workforce Development Agency Bill, this formation of SWDA will bring the skills and employment facilitation capabilities of SkillsFuture Singapore and Workforce Singapore under one roof, making it more seamless, more integrated for individuals and employers to obtain the appropriate support.
We agree with Mr Ng Chee Meng that the intelligence that we have in gleaning and harvesting all the data in the SWDA, this intelligence must continue to be built on a foundation of trust, and we look forward to working closely with our tripartite partners to ensure that our assessment of the labour market continues to be grounded and also current. This will be an important part of how we stay ahead of disruption and support workers driven by AI changes.
This includes platform workers facing the deployment of AVs, as Ms Yeo Wan Ling highlighted. MOM and SWDA, they are already working closely with the Ministry of Transport (MOT) and tripartite partners to strengthen transition pathways for these drivers, ahead of actual AV deployment. I want to add that actually, as a proxy, it is SWDA. But actually, it is SkillsFuture Singapore and Workforce Singapore working closely with MOT currently.
Second, we will do more to improve Singaporeans' AI literacy. Today, there are over 1,600 AI-related courses on the MySkillsFuture website. We will introduce diagnostic tools for individuals to assess their current level of AI readiness and find courses which suit their needs, deliver proven training outcomes that are aligned with employer demand.
From the second half of this year, Singaporeans who enrol in the selected SkillsFuture AI courses will receive six months of free access to premium AI tools. This will help them to apply classroom learning to their daily lives and work. Mr Kenneth Tiong suggested to make this access universal, without condition. That was something that the Government considered carefully. But not all Singaporeans require frontier agent-grade tools. For many, free versions are good enough and widely available.
By tying subsidies to training, we are better able to target those who are more serious about levelling up the use of AI and we help them to make optimal and responsible use of such powerful tools. As Assoc Prof Terence Ho and Mr Alex Yeo shared about earlier on, we hope Singaporeans will tap on the resources available and be proactive in their learning journey.
As Minister of State Jasmin Lau shared, IMDA will also be expanding the TechSkills Accelerator programme to develop AI-bilingual workers, starting with accountancy, legal and HR professionals. [ Please refer to " Clarification by Minister for Manpower ", Official Report, 6 May 2026, Vol 96, Issue 30, Correction By Written Statement section. ] More details will be shared in due course.
Third, to support businesses, I have also emphasised time and again that we have set aside over $400 million for the Enterprise Workforce Transformation Package. I do not want to go too much into it because I believe I have already covered in most of my Second Reading speech yesterday. But Mr Yip asked whether the grants could be tied to worker outcomes conditions. Today, businesses tapping on the Workforce Development Grant (Job Redesign+) , are required to support workforce outcomes, such as wage growth and retention as part of their transformation plan. Later this year, eligible businesses will also receive $10,000 under the redesigned SkillsFuture Enterprise Credit and this can be used to offset out-of-pocket costs for eligible workforce transformation programmes, including those under the Enterprise Workforce Transformation Package.
We agree with Mr Mark Lee that trade associations and Enterprise Workforce Transformation Package addresses these issues while ensuring that the workforce is brought along on the journey. Chambers play a particularly important role in connecting firms with the right expertise and resources. That is why we have appointed the Singapore Business Federation and SNEF as anchor programme partners for the Enterprise Workforce Transformation Package so that integrated workforce transformation support can be brought directly to firms, and we can help to accelerate AI adoption across sectors.
We are also supporting the Labour Movement's efforts to transform businesses and workers. The Government topped up the NTUC CTC Grant by around $200 million in 2025 and extended the grant to 2028. More recently, we worked with NTUC to expand the grant to better support Queen Bee companies to drive cluster-level transformation.
Notwithstanding the fact that I stepped out for a short while to answer a call, I am heartened to hear Mr Ng Chee Meng's sharing of how CTC has helped many businesses transform, while improving the lives of workers. I particularly note his exhortation and his suggestion to further expand the CTC initiative and share his ambition to elevate the CTCs to a tripartite level. We look forward to working with tripartite partners to jointly explore ways to make this a reality.
There are calls for us to go beyond project level interventions, to make a more structural shift in financial incentives for companies to invest in workers. Structural mechanisms, the likes of what Mr Andre Low called for, already exists. Grants, like the SkillsFuture Enterprise Credit, create direct financial incentives for companies to invest in the capabilities of their workers. We will continue to review and enhance such support as part of the work of the SWDA.
I think all of us should appreciate the importance of supporting enterprise transformation, not as a blanket, boiling the ocean strategy, but differentiated, precise and targeted sector-by-sector, company-by-company supporting their enterprise transformation. Even though it is more tedious, I believe it is also in the long run, more sustainable.
Lastly, we have strengthened transition support for workers who are displaced, so that they can bounce back stronger. The Government cannot protect every job but we will certainly do our best to support and protect every worker because every worker matters.
So, with AI transition, work processes will reorganise and change, jobs will also change and some jobs may get replaced. Going through transitions can be challenging. But I assure all of our workers that you will not walk alone.
We have recognised for some time that we must strengthen our support mechanisms as the pace of change accelerates. This is why we launched the SkillsFuture Jobseeker Support scheme (JSS) last year. This is as part of our refreshed social compact under Forward Singapore. The JSS provides temporary financial relief and job search support to involuntarily unemployed individuals. It has made a difference for many Singaporeans, helping them to regain their footing and to return to work with confidence.
Perhaps Mr Andre Low may have some misperception about the scheme. It is not a redundancy insurance because it does not just merely provide a cash payout for displacement. It is a support for re-employment. The JSS supports workers in their re-employment journey. It provides a degree of financial support for the lower- and middle-income precisely so that they do not rush into the first available job that may not be a good fit.
Workforce Singapore compliments the Jobseeker Support Scheme with hands on wraparound support to improve the quality of their job search. And we are cognisant of the fact that prolonged unemployment can harm a worker's longer-term career prospects and that is why the financial support is time bound. It tapers downwards because we believe that the first two to three months when the worker is involuntarily unemployed is when the impact is most felt. So, what we do is we try to raise the level in the initial part to encourage workers, to provide that lift and when it tapers downward, we hope that the workers will be able to find the right jobs for them.
But we hear calls. Mr Ng Chee Meng and Mr Patrick Tay proposed to raise the JS scheme income threshold to better support higher-income individuals. We will look at how the scheme can be improved and we will study this carefully.
We also hear Mr Ng Chee Meng's call for earlier notification of retrenchments to the Government, before employees' last working day, and Mr Mark Lee's reflection of businesses' concerns on this. We want to strike the right balance. Tripartite partners are already discussing shortening the retrenchment notification duration under the ongoing Employment Act review. We, on our part, would like to see notification to the Government happening before or by the last day of work of the affected workers as far as possible, because then this would also enable a timelier employment facilitation support to workers.
To Mr Kenneth Tiong's suggestion on strengthening protections for displaced workers, the Employment Act already provides broad-based protections by establishing procedural safeguards, like notice periods and dispute resolution avenues. This applies to all types of displacements, not just due to AI.
Our AI-enabled growth must be anchored in fairness, resilience and shared opportunity and this is not something that would happen naturally. Mr Vikram Nair asked what safeguards we have to ensure that workers are treated fairly as AI adoption increases. The Government has developed frameworks, such as Model AI Governance Framework for Agentic AI and AI Verify to establish clear responsibilities for actors across the AI supply chain, giving clarity to AI developers and users on responsible practices, including HR technology solution providers. And Mr Saktiandi Supaat rightly pointed out that AI adoption is uneven at varying speeds across sectors, worker segments and businesses of different sizes. Without deliberate effort, the gains from AI could flow to some while others are left behind. In China, the courts have ruled that it is illegal to replace employees with AI purely to cut costs.
Senior Minister of State Desmond Tan and Mr Sanjeev Kumar Tiwari spoke about the work NTUC has done in recent years to equip workers with AI-related skills and supporting workforce transformation. These are exactly the kind of capabilities we should draw on to ensure that more workers and businesses know what support is available and that AI adoption can accelerate across the economy. And this is why we supported wholeheartedly NTUC's proposal to form the Tripartite Jobs Council. The Tripartite Jobs Council will take a coordinated tripartite approach to mobilise enterprises and workers towards fair and resilient growth in an AI era.
As Assoc Prof Terence Ho noted, AI must augment workers, not replace them. We will leverage SNEF’s business advisors and NTUC's CTCs to help businesses adopt AI in ways that drive growth and enhance job roles, prioritising technologies that augment human capabilities and not replace them. We will harness our tripartite partners' strong links with workers, unions and employers to drive broad-based AI training across sectors and career stages, so that no worker is left behind as AI reshapes our journey.
In those technologies or businesses that have to undergo restructuring, we will work with the businesses to help pivot, upskill and reskill the workers.
We will also pay special attention to students and younger workers who are anxious about AI's impact on entry-level jobs. The IHLs continue to enhance their curriculum to keep pace with AI advancements. All IHLs will offer selected AI-related courses for their alumni at a significant discount for a year, starting in the second half of this year.
Graduates entering the workforce can also tap on MOE's SkillsFuture Work-Study Programmes, which combine classroom training with on-the-job training at companies to build both the skills and experience that employers value.
Assoc Prof Jamus Lim called for the expansion of youth apprenticeship pathways. We agree. Structured learning must be complemented with real workplace experience. We will continue to work with sector leads, like the Monetary Authority of Singapore and IMDA to support apprenticeships in high growth sectors, learning from our experience from programmes, like GRIT, we stand ready to refine and to expand these programmes, if necessary.
Mr Speaker, Sir, let me conclude. Singapore has weathered deep disruptions before, from the Asia Financial Crisis to SARS, to COVID-19. Each time, each crisis we came through not because the Government had all the answers, but because workers, businesses and Government stood shoulder-to-shoulder. That is the strength of tripartism.
In many countries, AI becomes a tug-of-war. Workers on one end, business on the other. Progress contested, trust strained. Singapore does not have to go down that road. We work together to make our entire economic pie bigger and make sure that the benefits are widely shared.
To our workers wondering where you stand, there will always be a place for you. Your experience, your judgement, matters more than ever and your commitment to our country, your support through the years, through the decades, we are deeply appreciative. Thank you very much. [ Applause. ]
To all of our young graduates, your ideas, your drive, matters more than ever. Your enthusiasm, your curiosity, that connect, that curiosity that Minister of State Jasmin Lau talked about just now, matters more than ever and we are behind you.
For all of our businesses, if you are unsure, you are uncertain as to where to start, you do not have to figure it out alone. We will walk alongside with you. We will help you to transform, will help you to compete so that together you can create better opportunities both for your businesses and for your workers.
We will not leave the future of work, the livelihoods of our workers, our Singaporeans, to chance. We will shape a transformation that is inclusive, forward thinking, anchored in real action.
Singaporeans will never be helpless passengers to an AI-driven future, but Singaporeans will be our fellow co-pilots as our AI journey takes flight. And we will move forward in the Singapore way with Government, employers and the unions working together to ensure that our AI transformation creates good jobs, clear pathways for every Singaporean worker towards a better future because every worker matters. With this, I rise in support of the Motion. [ Applause. ]
6.50 pm Mr Speaker : In Parliament, we are also adopting and embracing AI as we equip our staff on this journey. Mr Andre Low has a clarification? Leader, please move the exemption first.