The Federal Reserve’s top banking regulator just told the industry something it rarely hears from Washington: we’re not going to tell you exactly how to do this.
Fed Vice Chair for Supervision Michelle Bowman used a Bank Policy Institute conference in London on July 13 to lay out her vision for how regulators should handle artificial intelligence in finance. The short version: give banks guardrails, not a straitjacket.
A principles-based playbook for AI in banking
Bowman’s argument centers on four guiding principles she wants to anchor the Fed’s approach to regulatory modernization. First, focus on material financial risks, not every conceivable edge case. Second, tailor regulations to specific risk profiles rather than applying blanket rules across institutions of wildly different sizes and complexity. Third, maintain transparency and accountability in the supervisory process itself. Fourth, adopt a forward-looking posture that actually supports innovation instead of reflexively squashing it.
This wasn’t a one-off remark. Bowman first floated the concept in a May 1 speech, where she urged fellow regulators to reassess existing supervisory guidance to avoid piling excessive burdens on financial institutions. The London address doubled down on that message, framing it as part of a broader modernization push.
