MAS releases AI Governance Framework v2 for Singapore financial institutions — updated model risk management for generative AI, third-party AI vendor risk, and customer-facing AI disclosure requirements. Mandatory compliance expected within 18 months of final issuance.
The Monetary Authority of Singapore (MAS) has released Version 2 of its AI Governance Framework for financial institutions operating in Singapore — substantially updating the 2019 original to address the governance challenges specific to generative AI deployment, large language model-based decision systems, and third-party AI model provider risk that the 2019 framework did not contemplate.
MAS AI Governance Framework v2's principal additions to the original framework address four areas that MAS has identified as elevated risk in Singapore's financial sector AI deployment: generative AI model risk management (extending MAS's existing MRM Guidelines to cover LLM-based systems used in credit decisioning, fraud detection, and customer service); third-party AI vendor risk (requiring Singapore financial institutions to assess and document the AI model governance practices of AI API providers whose models are deployed in production financial services); customer-facing AI disclosure (requiring financial institutions to disclose to retail customers when AI systems materially influence financial product recommendations, credit decisions, or investment advice); and APAC data residency for AI inference (requiring that customer financial data used for AI inference remain within Singapore or MAS-designated data residency jurisdictions).
The framework is released in consultation form with an 8-week comment period — following which MAS will issue final guidance with implementation timelines. Industry commentary from Singapore Banking Association and FinTech Association of Singapore during the consultation will likely focus on the third-party AI vendor risk assessment requirements, which require financial institutions to obtain governance documentation from AI model providers (OpenAI, Anthropic, Google) that those providers are currently not structured to provide at the level of specificity MAS's framework contemplates.
For APAC financial technology companies and Singapore-based financial institutions with existing or planned AI deployments, MAS AI Governance Framework v2 represents the most significant Singapore AI compliance development since the original MRM Guidelines. APAC FinTech compliance teams should initiate gap assessments against the consultation document now rather than waiting for the final framework — the 18-month implementation window after final issuance begins from a known baseline, and APAC financial institutions that start late will face compressed timelines.
How AIMenta helps clients act on this
Where this story lands in our practice — explore the relevant service line and market.
Beyond this story
Cross-reference our practice depth.
News pieces sit on top of working capability. Browse the service pillars, industry verticals, and Asian markets where AIMenta turns these stories into engagements.
Other service pillars
By industry
Other Asian markets
Related stories
-
Regulation ·
Japan FSA Finalises AI Model Risk Management Framework for Financial Institutions
Japan's Financial Services Agency finalises AI model risk management framework requiring Japanese financial institutions to document model validation processes, report AI-related incidents within 48 hours, and conduct annual AI system audits — applying to AI-assisted credit scoring, algorithmic trading, fraud detection, and customer service AI deployed by Japanese banks, insurers, and securities firms.
-
Regulation ·
Singapore PDPC Issues Mandatory AI Impact Assessment Guidelines for Financial Institution AI Models
Singapore's PDPC issues mandatory AI impact assessment guidelines for financial institutions using AI in credit scoring and fraud detection — requiring documented bias evaluation, explainability reports, and quarterly senior management sign-off for high-risk AI models.
-
Research ·
NUS and MIT Research Shows APAC-Language LLMs Outperform English-First Models on Legal and Financial Reasoning
NUS and MIT publish multilingual LLM reasoning research showing APAC-language models trained on Mandarin and Japanese outperform English-first models on APAC legal and financial benchmarks by 18-31 percentage points.
-
Funding ·
Singapore AI Startup Imbue Raises $200M Series B for Autonomous Enterprise Agent Platform
Singapore AI startup Imbue raises $200M Series B to build autonomous AI agents for APAC enterprise workflows — targeting insurance claims processing, financial compliance automation, and supply chain decision-making in Singapore and Southeast Asian markets.
-
APAC ·
MAS Singapore Launches GenAI Regulatory Sandbox for APAC Financial Institutions
Singapore MAS launches GenAI sandbox for APAC financial institutions to test large language models in a regulatory environment. Gives APAC fintechs supervised access to trial GenAI in credit decisioning, fraud detection, and customer advisory without full regulatory approval.