Bowman advocates lighter oversight for lower-risk bank AI
Federal Reserve Vice Chair Michelle Bowman outlined sound practices for artificial intelligence at an FSB event on July 7, 2026. She called for proportional supervision that applies lighter oversight to lower-risk AI deployments across financial institutions.
Tiered controls for material deployments
The FSB consultation report on responsible AI adoption aims to establish international standards for financial institutions ahead of the G-20 deliverable later in 2026.
Bowman emphasized that risk management depends on understanding specific use cases and assessing whether AI applications are material to business operations or regulatory obligations.
Rather than imposing uniform mandates, the framework matches the intensity of governance and controls to actual risk profiles.
Under this approach, lower-risk deployments receive lighter supervisory oversight, while complex or higher-risk applications require more stringent safeguards.
The report includes case studies to demonstrate appropriate governance structures.
One size fits none in AI rules
The Federal Reserve has monitored bank usage of AI for nearly a decade, observing expanding adoption across institutions of all sizes.
The FSB workstream, led by the Monetary Authority of Singapore in collaboration with the US Treasury and SEC, prioritizes proportionality so smaller banks are not burdened by rules designed for complex institutions.
Regulators are actively soliciting public feedback to ensure the guidance avoids being overly prescriptive while addressing material risks effectively.
Pragmatic balance or regulatory illusion
Bowman rightly champions proportionality to stop compliance burdens from stifling bank innovation.
Yet drawing a clear boundary between low-risk and material AI deployments remains highly subjective.
Without precise supervisory benchmarks, this flexible framework risks offering vague protection against algorithmic failures.