⚖️ Governance Framework Statutory Agency Active Founded 2016-10

IMDA

Lead Ministry
Ministry of Communications and Information (MCI)
Scale / KPIs
1,500+ staff; oversees telecoms, broadcasting, media, and AI governance
Website
imda.gov.sg
Last Updated
2026-05-02

IMDA (Infocomm Media Development Authority) is Singapore's integrated regulator covering "infocomm + media + AI governance", formed in 2016 from the merger of IDA and MDA. On the AI front, it is **the central designer of Singapore's AI governance system** — virtually every major governance document, from the Model AI Governance Framework (MGF) to AI Verify and the Generative AI governance framework, originates from IMDA.

📖 What it is

IMDA is Singapore's core regulator for the digital economy, with responsibilities spanning:

  • Telecoms and internet regulation: spectrum, broadband, net neutrality, and other infrastructure oversight
  • Media regulation: broadcasting, film classification, content review
  • Digital transformation: pushing enterprise and government digitalisation (SGTech, SMEs Go Digital, etc.)
  • AI governance: setting national AI governance frameworks; the lead agency for exporting Singapore's "AI governance standards"

On AI governance, IMDA does not follow the EU AI Act's "mandatory legislation" path. Instead, it walks a "principles + voluntary adoption + internationalisation" route. The flagship outputs of this approach:

  • Model AI Governance Framework (MGF): first edition in 2019, v2 in 2020, then a Generative AI Framework in 2024
  • AI Verify: released in 2022, the world's first open-source AI governance testing framework (later spun off into the AI Verify Foundation)
  • AI governance international cooperation: jointly developing AI standards with OECD, GPAI, and ISO

🤖 Relation to AI

IMDA's role in AI is not "doing AI" but "defining how AI is done."

Its strategic choices are very clear:

  • Avoid the legislative route: side-step the EU AI Act's "heavy regulation, slow deployment" trap
  • Take the "tooled governance" route: turn governance principles into runnable test suites (AI Verify) so enterprises can self-assess
  • Anchor to international standards: MGF deliberately aligns with NIST AI RMF, ISO/IEC 42001, and OECD AI Principles, so IMDA's standards gain international recognition

The advantage of this path: Singapore can export AI governance "tool standards" without sustaining a large regulatory bureaucracy or passing legislation. AI Verify has been adopted by 50+ enterprises globally (including IBM, Singtel, Standard Chartered) — a real expression of Singapore's soft power.

The downside is just as obvious: voluntary adoption = no teeth. When an AI system causes real harm, IMDA lacks direct enforcement tools and has to lean on sector regulators like PDPC, MAS, and MOH to follow up.

🇸🇬 Relation to Singapore

IMDA is the key hub for Singapore's externally projected AI capability.

In the "seven transmission levers" framework:

  • Lever 4 (governance): IMDA is the chief designer of Singapore's AI governance
  • Lever 6 (international): MGF and AI Verify are Singapore's "governance calling cards" at GPAI, the Bletchley/Seoul summits, and the OECD
  • Lever 3 (industry adoption): programmes like SMEs Go Digital push enterprise AI deployment

A take: IMDA's AI governance route is a textbook example of Singapore's "small country, big strategy" — instead of competing with the EU on legislation, the US on technology, or China on scale, it stakes out the "governance tools and standards" niche. AI Verify was a particularly clever move: open-source, internationally usable, and tooled — making Singapore the "Switzerland of AI governance".

But IMDA also has structural challenges: with telecoms, media, and AI all under one roof, attention is badly diluted. AI governance is its "third child" — competing with telecoms regulation for budget, attention, and talent. Whether AI governance functions should be spun off in the NAIS 2.0 era (e.g., handed over to the AI Verify Foundation) is a question worth watching.

🗓️ Key Milestones

  1. 2016-10
    IDA + MDA merge to form IMDA
  2. 2019-01
    Released Model AI Governance Framework v1

    Among the first national-level AI governance frameworks worldwide.

  3. 2020-01
    MGF v2 released

    Added implementation case studies and self-assessment checklists.

  4. 2022-05
    AI Verify testing framework released

    World's first open-source AI governance testing suite.

  5. 2023-06
    AI Verify Foundation spun off
  6. 2024-05
    Model AI Governance Framework for GenAI released

👥 Key People

🔗 Related

Sources

Within ⚖️ Governance Framework