MAS Mandates Independent AI Reviews for Singaporean Financial Firms
AI-generated image · US National Wire
New regulatory guidelines shift the burden of AI liability onto fintechs, requiring third-party audits and strict vendor oversight.
Singapore’s Monetary Authority (MAS) is signaling that the efficiency gains of artificial intelligence will not exempt financial institutions from liability. In its first Guidelines on Artificial Intelligence Risk Management for Financial Institutions, MAS stipulates that all AI use cases must undergo independent review by parties not involved in their development before they can be deployed into production.
According to reporting from The Register, the regulator warns that the probabilistic nature of AI—and the even greater complexity of generative and agentic AI—creates uncertainty and risks of biased or unexpected behavior. To mitigate this, MAS requires firms to conduct technology and cybersecurity reviews to ensure controlled deployment.
Crucially, MAS is closing the loophole on vendor liability. The guidelines state that financial institutions remain accountable for the AI services they deliver, even when those systems are developed or operated by third parties. The regulator expects firms to obtain sufficient assurance from providers and implement compensating controls to fill assurance gaps. If risks exceed a firm's risk appetite, MAS suggests the service be limited, suspended, or replaced.
Beyond initial audits, the guidelines mandate ongoing monitoring to combat model staleness and performance degradation. Firms must also maintain up-to-date inventories of all AI in use and develop contingency plans—including manual processes or alternative systems—for high-risk applications to ensure business continuity during AI failures.
The guidelines are set to take effect on October 7, 2027.