The Monetary Authority of Singapore (MAS) has published a consultation paper outlining proposed Guidelines on Artificial Intelligence (AI) Risk Management which will apply to all financial institutions (FIs) in Singapore.
Under the guidelines, boards and senior management must assume direct accountability for AI governance and oversight. If an institution’s overall AI risk exposure is material, a dedicated cross-functional committee may be required.
Financial institutions will be required to maintain an accurate inventory of all AI use cases and to conduct a Risk Materiality Assessment that considers dimensions such as impact, complexity and reliance on the AI system.
Lifecycle controls are central: the guidelines cover data-management, fairness, transparency, explainability, human oversight, third-party risk, model evaluation and monitoring, and change management. These controls are expected to be applied proportionately relative to the institution’s AI exposure.
The guidelines cover a broad spectrum of AI technologies including generative AI and emerging AI agents, signalling MAS’s intent to regulate not just legacy AI applications but frontier systems as well.
MAS has invited industry feedback, with the consultation open until 31 January 2026; a transition period of 12 months is expected following finalisation of the guidelines.
For insurers, banks and other financial-services firms operating in or interacting with Singapore’s ecosystem, the proposed guidelines signal a significant escalation in expectations around AI risk management. Firms will need to revisit their vendor-AI risk exposure, model-governance frameworks, human-in-the-loop processes and audit trails to align with MAS’s supervisory expectations.
For more details and structured learning, please explore our risk management courses.
