MAS issues draft AI risk guidelines for all financial institutions

Guidelines cover governance, lifecycle controls and proportional AI safeguards

The Monetary Authority of Singapore (MAS) issued a consultation paper proposing guidelines on AI Risk Management for all financial institutions (FIs), outlining supervisory expectations for the use of artificial intelligence.

The guidelines apply across the financial sector and cover governance, risk systems, lifecycle controls, and institutional capabilities. The scope includes Generative AI and emerging technologies such as AI agents.

MAS stated that requirements should be applied in proportion to each FI’s AI usage, business model, and risk profile. 

Key expectations include board-level oversight, firm-wide AI inventories, and risk materiality assessments. Lifecycle controls must address data governance, fairness, transparency, human oversight, third-party risks, and change management, scaled to the level of risk.

The guidelines follow MAS’ 2024 thematic review of AI use in key banks and industry consultations. MAS said the framework sets out supervisory standards for managing AI-related risks.