Singapore AI Governance Cheat Sheet — Free reference

Key laws

PDPA Section 11: accountability obligation (DPO appointment, policies)
PDPA Section 13: consent obligation
PDPA Section 14: notification of purposes obligation
PDPA Section 16: withdrawal of consent / purpose limitation
SG Model AI Governance Dimension 1: internal governance structures and measures
SG Model AI Governance Dimension 2: determining the level of human involvement in AI-augmented decision-making
SG Model AI Governance Dimension 3: operations management (data quality, model training, monitoring)
SG Model AI Governance Dimension 4: stakeholder interaction and communication

AI Verify (IMDA) — confirm your AI systems are tested against the framework's 11 ethical principles. The 2025 toolkit update added GenAI-specific evaluation modules.
PDPA — Singapore Personal Data Protection Commission's AI advisory guidelines (2024) require DPIAs for AI training on personal data. Confirm yours are current.
MAS FEAT principles (Fairness, Ethics, Accountability, Transparency) — financial-services AI must demonstrate compliance via the FEAT Self-Assessment Toolkit.
Model AI Governance Framework (PDPC, 2nd edition 2020) still the foundational reference. The 2024 Generative AI Governance Framework supplements it for foundation models.
Healthcare AI: MOH's AI in Healthcare Guidelines (2024) require clinical validation reporting. Confirm your AI clinical decision support has the validation evidence pack.

One-page .docx · ~10 KB · free. You'll also be subscribed to the weekly Brief.