Singapore's AI Policy in 2026: IMDA, the Model AI Governance Framework, and the Strategic Position
Singapore's AI policy approach has been deliberately pragmatic. Where the framework sits in 2026 and what enterprises should know.
Singapore’s AI policy approach has been deliberately pragmatic — voluntary frameworks, sector-specific guidance, and strong international coordination, rather than horizontal prescriptive law in the EU AI Act style. The Model AI Governance Framework, AI Verify (the open-source testing toolkit), and the AI Safety Institute (Singapore) collectively constitute the country’s AI governance architecture. By 2026 the framework has produced substantial international influence.
I want to walk through where Singapore’s AI governance actually sits.
The framework#
The Model AI Governance Framework is the foundational policy document, originally issued 2019 and updated multiple times. The framework is voluntary but operationally influential. It articulates principles for responsible AI use, organizational practices, and decision-making considerations.
AI Verify is the open-source AI governance testing framework — toolkit and best-practice library for organizations to assess and demonstrate the trustworthiness of AI systems. Increasingly adopted internationally as a reference framework.
AI Verify Foundation is the multilateral body coordinating AI Verify development and adoption.
Singapore AI Safety Institute is part of the international AISI network coordinating frontier-AI safety evaluation.
Sector regulators — MAS for financial services, MOH for healthcare, MOM for employment — issue sector-specific AI guidance.
What enterprises should know#
Singapore’s approach is:
-
Voluntary alignment with the Model AI Governance Framework is increasingly expected.
-
AI Verify testing is increasingly used in procurement and certification.
-
Sector-specific compliance is the primary regulatory work.
-
Privacy under PDPA (covered here) applies to AI processing personal data.
-
International alignment with AISI evaluations, ISO standards, and other frameworks.
How Singapore compares#
Singapore’s pragmatic approach contrasts with:
- EU AI Act — prescriptive, horizontal, risk-based with mandatory conformity.
- US sector-and-state patchwork.
- Japan’s sector-led approach (covered here) which is the most-similar to Singapore’s.
- UK’s pragmatic approach which is similar in spirit.
Singapore’s framework has been particularly influential in shaping how other Asia-Pacific countries approach AI governance.
What’s coming in 2026 and 2027#
Three things to watch:
AI Verify continued development with broader international adoption.
Sector-specific guidance continues to be issued.
The AISI Singapore role in international cooperation expands.
Where pdpspectra fits#
Our AI engineering and compliance work spans Singapore and the broader Asia-Pacific. We work with enterprises on AI governance, evaluation, and deployment.
Related reading: the Japan AI policy post, the UK AI Safety Institute post, and the EU AI Act post.
Singapore’s AI policy is pragmatic and influential. Talk to our team about your AI deployment.