Consultation Paper on Proposed Guidelines on Artificial Intelligence Risk Management for Financial Institutions (P017-2025)
Is this legally binding?
Consultation. Proposed, not final. Read it, plan for it, and do not treat it as settled.
Proposed and not final. Nothing here can be enforced against you yet.
MAS consultation paper (published 13 Nov 2025) proposing Guidelines on AI Risk Management for all financial institutions: AI oversight, risk management systems/policies/procedures, AI life cycle controls, capabilities and capacity. Closed 31 Jan 2026. AS OF 2026-08-17 STILL PROPOSED, NOT FINAL: consultation page shows 'Closed' with no response to feedback and no final Guidelines.
From the source
“This consultation sets out MAS’ proposed Guidelines on Artificial Intelligence (AI) Risk Management.”
MAS consultation page P017-2025; Consultation Paper Preface
Instrument record
- Short name
- MAS AI Risk Consultation
- Date
- 2025-11-13
- Version
- Not versioned
- Cluster
- core
What this connects to
47 relations. Official relations are the ones the source documents state; anything marked GAGE analysis is our reading, not an agency's.
Operationalised by1
- InstrumentMindForge AI Risk Management Toolkit (Project MindForge phase 2)Guidance
Handbook 'intended to accompany and support the implementation of the proposed MAS Guidelines on AI Risk Management'; its four sections are aligned with the proposed Guidelines (MR para 3)
Issued by1
- AgencyMonetary Authority of Singapore (MAS)Binding, sectoral
Consultation paper published by MAS
Applies to1
- ConceptFinancial services sector (Singapore)Guidance
Proposed to apply to all FIs as defined in s.2 FSM Act 2022
Cites6
- InstrumentPrinciples to Promote Fairness, Ethics, Accountability and Transparency (FEAT) in the Use of AI and Data Analytics in Singapore's Financial SectorSupervisory expectation
CP para 2.4 and footnote: the Guidelines complement MAS' FEAT principles, which continue to apply
- InstrumentVeritas Initiative (MAS-led consortium)Guidance
CP para 2.4: Veritas Initiative supports FIs in incorporating FEAT into AIDA solutions
- InstrumentModel AI Governance Framework, 2nd Edition (2020)Guidance
CP para 2.7/footnote 9: Guidelines complement national initiatives such as IMDA's Model AI Governance Framework
- InstrumentAI Verify Testing Framework and ToolkitGuidance
CP para 2.7/footnote 10: Guidelines complement the initiatives under the AI Verify Foundation
- InstrumentMAS Notices on Technology Risk Management (TRM Notices)Binding, sectoral
Proposed Guidelines paras 4.16/5.3 (footnotes 28, 30): existing MAS technology risk management guidelines and notices apply to AI systems
- External frameworkNIST AI Risk Management Framework 1.0 (extern crosswalk target)Guidance
Proposed Guidelines para 5.3 footnote 31 references the NIST AI Risk Management Framework for technology infrastructure
Cited by31
- EventP017-2025 consultation opened (13 Nov 2025)Consultation
Consultation milestone
- EventP017-2025 consultation closed (31 Jan 2026); still not finalised as of 2026-08-17Consultation
Consultation milestone
- SectionApplicability and proportionality (Consultation Paper Section 3; proposed Guidelines Section 1)Consultation
Section of the consultation paper / proposed Guidelines
- SectionProposed Guidelines Section 2: AI oversightConsultation
Section of the consultation paper / proposed Guidelines
- SectionProposed Guidelines Section 3: key AI risk management systems, policies and proceduresConsultation
Section of the consultation paper / proposed Guidelines
- SectionProposed Guidelines Section 4: key AI life cycle controlsConsultation
Section of the consultation paper / proposed Guidelines
- SectionProposed Guidelines Section 5: AI capability and capacityConsultation
Section of the consultation paper / proposed Guidelines
- SectionAnnex: proportionate application of the proposed GuidelinesConsultation
Section of the consultation paper / proposed Guidelines
- ConceptProposed scope of AI: model, system, use caseConsultation
Concept defined in the consultation paper
- ConceptRisk materiality dimensions: impact, complexity, relianceConsultation
Concept defined in the consultation paper
- ConceptProposed 12-month transition periodConsultation
Concept defined in the consultation paper
- ObligationProposed: board senior mgmt oversightConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: dedicated ai committeeConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: basic ai policiesConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: ai identificationConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: ai inventoryConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: risk materiality assessmentConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: data managementConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: transparency explainabilityConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: fairness controlsConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: human oversightConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: third party ai managementConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: selection evaluation testingConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: technology cybersecurityConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: reproducibility auditabilityConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: pre deployment reviewsConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: post deployment monitoringConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: incident management kill switchConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: change management decommissioningConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: capabilities trainingConsultation
Proposed expectation in P017-2025 (not final)
- ObligationProposed: technology infrastructureConsultation
Proposed expectation in P017-2025 (not final)
Maps across to6
- External frameworkClause 5: LeadershipStandardGAGE analysis, not official
Proposed MAS AI risk management expectations emphasise board/senior-management accountability per AIMS leadership
- External frameworkFunction: GOVERNGuidanceGAGE analysis, not official
MAS consultation proposals map closely to the GOVERN function for FIs
- External frameworkArticle 9: Risk management systemBindingGAGE analysis, not official
MAS AI risk management proposals parallel the Article 9 risk management system
- External frameworkArticle 17: Quality management systemBindingGAGE analysis, not official
MAS proposals on FI governance parallel provider QMS expectations
- External frameworkClause 6: PlanningStandardGAGE analysis, not official
MAS AI risk management proposals align with AIMS risk planning
- External frameworkClause 9: Performance evaluationStandardGAGE analysis, not official
MAS proposals on monitoring and review align with AIMS performance evaluation
Mapped across from1
- ConceptRisk-based classification: high, medium, lowBindingGAGE analysis, not official
Contrast: Vietnam's binding tiers vs MAS's sectoral supervisory risk approach
Verified against the official source on 2026-08-17. GAGE is not affiliated with or endorsed by any agency named here, and nothing on this page is legal advice. How this is built and checked.