Regulation

MAS finalises AI risk guidelines for financial institutions and plans an agentic AI consultation in 2027

Singapore's central bank and financial regulator has issued final Guidelines on AI Risk Management that apply to all forms of AI, take effect on 7 October 2027, and flag greater use of autonomous agentic systems as a reason to keep controls under review.

· Source: Monetary Authority of Singapore


On 7 October 2026 the Monetary Authority of Singapore (MAS) issued its final Guidelines on Artificial Intelligence Risk Management, following a public consultation in November 2025. The release says the guidelines apply to all financial institutions and all forms of AI technology, including AI models and systems with greater autonomy in generating outputs, decision making or execution. It also notes that the Financial Stability Board has separately consulted on sound practices for responsible AI adoption. MAS says it kept the key expectations from the consulted draft and refined them after respondents asked whether existing governance structures could be used, how to treat risks from embedded AI, and when basic governance policies are enough.

The expectations fall into four groups. Boards and senior management should oversee AI risks with clear accountabilities, and existing governance structures may be used where they give adequate oversight, so a dedicated AI committee is not required. Institutions should keep inventories of their AI use, assess how material each use case is, and apply proportionate controls across the AI life cycle, including data governance, testing, human oversight, cybersecurity, monitoring and change management. They remain accountable for AI supplied by third parties and should consider limiting, suspending or replacing a third-party AI service if its risks cannot be brought within their risk appetite. And a firm may use basic policies where poor performance or unavailability of an AI tool is unlikely to have a material impact. The guidelines take effect on 7 October 2027, with Sections 3 and 4 expected from that date and Sections 5 and 6 by 7 October 2028.

On agents, the release is specific but limited. It says institutions should review their controls as AI use expands and technology evolves, giving as an example the greater use of agentic AI systems that can operate autonomously and access tools, and that in 2027 MAS intends to consult the financial sector further on what additional guidance on agentic AI would be useful. So there is no agent-specific rule yet. For readers running AI trading agents through a broker or platform, the guidelines bind the regulated institutions in Singapore, not individual users, and the release does not mention trading agents. The practical read-across is about what those institutions will be expected to show: an inventory of AI in use, human oversight, testing and monitoring, and assurance from the third parties whose AI they rely on. What to watch next: the 2027 agentic AI consultation and whether other supervisors take a similar principles-based, proportionate approach.


Read the original: Monetary Authority of Singapore - MAS finalises AI risk guidelines for financial institutions and plans an agentic AI consultation in 2027. Commentary is the independent editorial view of Share Trading; the original article is credited to its publisher.