Skip to main content
Submit RFP

      The Monetary Authority of Singapore (MAS) has released a consultation paper on the proposed Guidelines on AI Risk Management (AIRG). These Guidelines introduce formal requirements focused on strengthening AI risk management across the entire AI lifecycle which is proposed to apply to all Financial Institutions. These developments are part of a national push to ensure proactive, proportionate risk management as the adoption of complex AI, including Generative AI and AI agents, becomes pervasive in the financial sector.

      Key expectations under MAS’s proposed AIRG

      find_replace

      Elevated Oversight

      Establishing clear Board and Senior Management responsibilities, including the potential requirement for a dedicated cross-functional committee if overall AI risk exposure is deemed material.

      task

      Structured Risk Systems

      Implementing robust systems for AI Identification, maintaining a comprehensive AI Inventory, and conducting consistent Risk Materiality Assessments across the dimensions of Impact, Complexity, and Reliance.

      tenancy

      Life Cycle Controls

      Applying proportionate controls throughout the AI life cycle in critical areas such as Data Management, Fairness, Transparency and Explainability, Human Oversight, and Third-Party AI Management.

      How KPMG can support your AIRG readiness

      MAS is currently running a public consultation, which ends on 31 January 2026. MAS also proposes a transition period of 12 months from the AIRG issuance date to implement these Guidelines. Given the breadth of the requirements, comprehensive planning and framework enhancements are essential now.

      KPMG has thoroughly analysed the Consultation Paper and worked closely with FIs on similar regulatory transitions. We would be delighted to support your compliance efforts and explore the following topics further in a dedicated briefing session:

      • Interpreting the proportionate application of the Guidelines.

      • Establishing the mandate and structure for the AI Risk Oversight Committee.

      • Developing and integrating the required AI Inventory and Risk Materiality Assessment methodologies.

      • Practical steps for ensuring compliance with AI Life Cycle Controls for both internally developed and third-party AI solutions.

      We would appreciate the opportunity to schedule a concise meeting at your earliest convenience to discuss how these Guidelines will impact your organisation and how we can assist in developing a robust compliance roadmap.