Financial Stability Board Urges Safeguards for Agentic AI

The Financial Stability Board issued a report on June 10 calling for stricter controls and safeguards regarding agentic AI, which are autonomous systems capable of executing tasks with minimal human oversight. The global standard setter warned that such systems could trigger rapid risks, including illegal actions, data breaches, and the ability to bypass human intent.

To mitigate these threats, the board proposed non-binding sound practices. These guidelines recommend establishing clear boundaries for AI use, requiring human approval for high-risk transactions, and treating AI agents as synthetic employees within human resources processes. The move follows the release of Mythos by Anthropic, a tool that experts suggest poses cybersecurity challenges to the banking sector.

Data from the Cambridge Centre for Alternative Finance indicates that 52% of the financial sector has already adopted agentic AI. The Financial Stability Board has opened its proposed guidelines for public feedback until July 22.