MAS issues final Guidelines on AI Risk Management for Financial Institutions — binding supervisory expectations across Singapore financial sector

On 7 October 2026 the Monetary Authority of Singapore issued final 'Guidelines on Artificial Intelligence (AI) Risk Management' for all financial institutions (FIs), following its November 2025 consultation. The Guidelines set out supervisory expectations for board/senior-management oversight of AI risk with clear accountability, risk identification and management across the AI lifecycle (data governance, testing, human oversight, cyber, monitoring, change management), third-party AI management, and risk-proportionate application. MAS notes it intends to consult in 2027 on further agentic-AI guidance. The Guidelines take effect 7 Oct 2027 (Sections 3-4) and 7 Oct 2028 (Sections 5-6).

This is a comprehensive binding AI risk-management framework for a financial sector applying to ALL financial institutions and ALL forms of AI including generative and agentic AI, making Singapore the first-mover setting explicit supervisory expectations for agentic finance. Any FI operating in Singapore — local or foreign — must align governance, inventories, testing, human oversight and third-party assurance to MAS expectations, and the 2027 agentic-AI follow-on signals more requirements on autonomous models.

FIs should map current AI use cases to MAS expectations, stand up AI inventories and lifecycle controls, review third-party AI vendor assurance, and plan phased implementation to meet the Oct 2027/2028 effective dates; monitor the 2027 agentic-AI consultation.