Guidelines  ·  2026-08-08

MAS (Singapore) confirms forthcoming AI Risk Management Guidelines will formally cover agentic AI

GuidelinesMedium impactSingapore
In a written parliamentary reply issued for the August 5, 2026 sitting, MAS Chairman/Deputy Prime Minister Gan Kim Yong confirmed that MAS's proposed Guidelines on Artificial Intelligence Risk Management (published for consultation in November 2025) will apply to all AI use cases at financial institutions, explicitly including agentic AI, and 'will be finalised soon.' MAS did not commit to making the industry-developed Safeguards for Agentic Finance at Runtime (SAFR) framework mandatory, nor give a finalization date. The guidelines themselves remain in pre-final/consultation status as of this reporting window — not yet a finalized publication.
This is the clearest confirmation yet from a major financial regulator that supervisory AI risk-management expectations will explicitly extend to autonomous/agentic AI systems, ahead of comparable moves by other regulators (EU AI Act agentic-specific rules not due until 2027). It signals imminent, sector-wide binding guidance for Singapore's 160+ licensed financial institutions, but as of this window it is a reaffirmation of scope via parliamentary reply, not the final guideline text itself.
Financial institutions operating in Singapore should begin mapping agentic AI deployments against the November 2025 consultation draft and the SAFR framework ahead of formal finalization; monitor MAS for the final Guidelines publication.
Fintech News SingaporeMAS written parliamentary replyOpenGov Asia
See this in the live feed Explore related AI security and governance findings — updated every morning.
Open the feed →