The Monetary Authority of Singapore released its Consultation Paper on Guidelines on Artificial Intelligence Risk Management on November 17, 2025 with consultation closing January 31, 2026. Final AIRM Guidelines remain pending as of July 2026 with a 12-month transition period to follow publication. MAS Chief FinTech Officer Kenneth Gay and Managing Director Chia Der Jiun have identified responsible AI adoption as a transformative priority. MAS published the AI Risk Management Toolkit in March 2026 as Phase 2 of Project MindForge covering traditional generative and agentic AI risk management. The AIRM Guidelines' four supervisory expectation areas - board oversight, AI risk management, AI lifecycle management, and third-party risk - do not specify the cryptographic audit trail required to prove in a Singapore court that a specific AI-governed financial decision was made by an authorised system under known constitutional parameters. This paper documents the structural gap between MAS's supervisory-ready AI governance framework and the constitutional command layer required to make that framework legally defensible in customer legal challenges.
Read paper: https://doi.org/10.5281/zenodo.21100290
For full paper and sovereign briefings: protocol@fijishi.com