🇸🇬 Singapore AI Observatory Dashboard

Data updated: 2026-05-02

Six dimensions on Singapore AI: strong investment, governance, and infrastructure; low talent self-sufficiency and limited frontier research are structural weaknesses. Numbers and editorial interpretation are shown side by side — readers decide.

International Anchors

Tortoise Global AI Index 2024 · #3 Oxford Government AI Readiness 2024 · #2 Microsoft AI Adoption 2026 · #2 (60.9%) WIPO Global Innovation Index 2025 · #5
💰

Investment Intensity

Is the government willing to spend?

↗
S$139 per person
vs US $33 / China $7 (per capita)
Government AI commitments > S$2B (NAIS 2.0 + Public AI R&D 2026–2030 + ECI); Budget 2026 adds 400% tax incentive + S$1.5B FSDF
Stanford AI Index 2025 · Per-capita government AI spend among global top
👩‍💻

Talent Pipeline

Are there enough people, and how self-sufficient is the supply?

↗
5,000 / 15,000
33% of 2029 target (35% foreign)
Tortoise Global AI Index — Talent sub-score · ~#6–8
🖥️

Compute Stack

Can it run frontier models?

→
1.4 GW
Data centre capacity + 70+ facilities + NSCC ASPIRE 2A+ at 20 PFLOPS
300MW additional capacity allocated + 80MW pilot 2026–2028 (incremental supply on the way, but power is the ceiling)
Tortoise Global AI Index Infrastructure · #2
🏢

Industry Adoption

Are enterprises actually using AI?

↗
62.5% large enterprises / 14.5% SMEs
SME adoption tripled YoY (4.2% in 2023 → 14.5% in 2024)
NAIIP target: 10K firms + 100K workers (2026–2029)
Microsoft AI Economy Institute 2026 · #2 globally (60.9%, behind UAE only)
🔬

Research Quality

Is original research coming out?

→
Per-capita papers #1 globally
NTU AI #3 (after MIT/CMU) · NUS AI #9
SEA-LION v4 (11 languages, 4B–33B params) + 100E (100+ projects) + ICLR 2025 hosted
Wiley 2024 · #1 globally for AI papers per capita (250 per million people, 2022)
🌐

Governance Influence

Is Singapore writing the rules?

↗
Rule-maker

Singapore Consensus on AI Safety signed by 11 countries (incl. US and China); ASEAN Guide on AI Governance adopted by all 10 ASEAN states (drafted under Singapore's lead); AI Verify Foundation cited globally; REAIM co-hosted; ISESEA held twice — Singapore is a rule-maker, not a rule-taker, with influence well above its size. Full participation in Bletchley, Seoul, and Paris AI Safety Summits; MAS Project MindForge has 24 institutions + the four major cloud vendors; UN Independent International Scientific Panel includes Singapore.

Oxford Government AI Readiness 2024 · #2 (behind the US only)

Method: Each dimension shows headline numbers + third-party rankings + target progress + trend + editorial interpretation — no overall grade. Full methodology →