AIPolicyTracker
Menu
Singapore Framework Voluntary standard Non-binding

Singapore Model AI Governance Framework: requirements, deadlines and compliance actions

Model AI Governance Framework (Second Edition) and Model AI Governance Framework for Generative AI

Source-linked Open official source JSON record

What is the Singapore Model AI Governance Framework?

Singapore's Model AI Governance Framework is a voluntary, sector-agnostic guide for organisations deploying AI. The second edition (January 2020) covers four areas: internal governance structures and measures, determining the level of human involvement in AI-augmented decision-making, operations management (data, model development, monitoring), and stakeholder interaction and communication. The May 2024 Model AI Governance Framework for Generative AI extends it with nine dimensions including accountability, data, trusted development and deployment, incident reporting, testing and assurance, security, content provenance, safety and alignment research, and AI for the public good.

Who does it apply to?

Voluntary; applicable to any organisation developing or deploying AI in Singapore or wishing to align with Singapore's expectations. It is referenced by the PDPC and IMDA and underpins the AI Verify testing framework.

Organisations deploying AI (any sector), and developers of generative AI models and applications.

When do the requirements apply?

First edition January 2019; second edition 21 January 2020; Generative AI framework 30 May 2024.

What must organisations do?

Nothing mandatory. Adopters set up governance structures, choose a human-in/over/out-of-the-loop model based on risk, manage data quality and model lifecycle, and communicate with users, with additional generative-AI measures.

Voluntary Establish internal governance structures and measures for AI Second edition, Part on internal governance structures and measures

Organisations should adapt existing governance to AI: clear roles and responsibilities, board and senior management oversight, risk-management and internal controls, and staff training.

Practical action: Assign an AI governance owner and add AI risks to the enterprise risk register.

Evidence examples: AI governance structure document

Framework mapping (original, editorial): ISO/IEC 42001:2023 Clause 5 Leadership; NIST AI RMF 1.0 GOVERN 2.x

Obligation page Source-linked

Voluntary Determine the appropriate level of human involvement in AI decisions Second edition, Part on human involvement in AI-augmented decision-making

Using a risk-impact matrix (probability and severity of harm), organisations choose human-in-the-loop, human-over-the-loop or human-out-of-the-loop designs and document the rationale.

Practical action: Record the human-involvement choice and rationale for each AI decision point.

Evidence examples: Risk-impact assessment and human-involvement decision

Framework mapping (original, editorial): NIST AI RMF 1.0 GOVERN 3.2, MANAGE 2.x

Obligation page Source-linked

Voluntary Manage data quality, model development and monitoring across the lifecycle Second edition, Part on operations management

Covers data lineage and quality, minimising bias in datasets, model explainability, repeatability, robustness, regular tuning and active monitoring after deployment.

Practical action: Maintain data lineage records and a model monitoring plan.

Evidence examples: Data lineage and model monitoring records

Framework mapping (original, editorial): ISO/IEC 42001:2023 Clause 8 Operation; Annex A data controls

Obligation page Source-linked

Voluntary Report incidents and mark AI-generated content (generative AI framework) Generative AI framework, dimensions on incident reporting and content provenance

The generative-AI framework recommends incident-reporting channels and processes for AI harms, and content provenance measures such as digital watermarking and cryptographic provenance so that users can identify AI-generated content.

Practical action: Implement provenance markers for generated media and an AI incident reporting channel.

Evidence examples: Content provenance implementation record

Framework mapping (original, editorial): NIST AI RMF 1.0 NIST AI 600-1 content provenance and incident disclosure

Obligation page Source-linked

Penalties

None; voluntary.

Official sources

  1. Model Artificial Intelligence Governance Framework (Second Edition)
    PDPC and IMDA · 21 Jan 2020 · Tier 1 source
  2. Model AI Governance Framework for Generative AI
    AI Verify Foundation and IMDA · 30 May 2024 · Tier 1 source
  3. AI Verify Foundation
    AI Verify Foundation · Tier 1 source

Change history

Record version 1: Initial structured record.. Full edit history is in the GitHub repository.

Frequently asked questions

Is Singapore's Model AI Governance Framework mandatory?
No. It is voluntary guidance from IMDA and the PDPC, but it reflects regulator expectations and underpins the AI Verify testing framework.

Informational only, not legal advice. Verify every claim against the linked official sources and consult qualified counsel before acting.