WASHINGTON—Michelle Bowman said the rapid emergence of artificial intelligence—particularly tools capable of identifying cyber vulnerabilities—is forcing regulators to rethink how financial institutions manage risk, even as the technology promises major efficiency gains across banking.
Speaking May 1 at a Financial Stability Oversight Council roundtable on AI and cybersecurity in Washington, the vice chair for supervision at the Federal Reserve Board said AI is quickly becoming a “force multiplier” for banks of all sizes, improving speed, modeling and operations, while also introducing new risks that require closer supervisory attention.
Bowman emphasized that regulators are shifting their supervisory approach to focus more narrowly on material financial risks, including how AI is deployed in critical functions such as credit decisions or enterprise-wide systems. She said supervisors are working to ensure guidance remains flexible enough to allow innovation—particularly for smaller institutions that may rely more heavily on third-party vendors—while still maintaining safety and soundness.
A key development highlighted in her remarks is the Federal Reserve’s recent move, alongside the OCC and FDIC, to revise model risk management guidance so it no longer broadly applies to newer forms of generative or agentic AI. Bowman said existing frameworks had been stretched beyond their original intent and could unintentionally hinder adoption of emerging technologies, adding that regulators are also working to simplify third-party risk guidance tied to vendor-provided AI tools.
At the same time, Bowman pointed to growing concern around advanced models such as Anthropic’s “Mythos,” which can rapidly identify system vulnerabilities. While such tools could strengthen cybersecurity defenses, she warned they could also be used maliciously to exploit weaknesses, underscoring the need for coordinated government-industry engagement and more dynamic supervisory responses.
Looking ahead, Bowman said U.S. regulators are coordinating with international bodies, including the Financial Stability Board, to develop consistent global approaches to AI oversight. A consultation report on AI adoption and risk is expected later this year, as regulators seek industry input on how to balance innovation with financial stability in an increasingly AI-driven environment.