WASHINGTON — Federal Reserve Vice Chairman for Supervision Michelle W. Bowman said Friday that rapid advances in artificial intelligence are underscoring the need for regulators to adapt supervisory approaches to keep pace with emerging technologies.
Speaking at an event hosted by the Financial Stability Oversight Council in Washington, Bowman pointed to new developments such as Anthropic’s Mythos model as evidence of how quickly AI capabilities are evolving.
Bowman said the model’s ability to detect cyber vulnerabilities illustrates both the promise and risk of advanced AI, noting it can be used to strengthen cybersecurity defenses or potentially for malicious purposes.
“As we learn more about this tool and others to be released in the coming weeks and months, we will continue to consider effective supervisory approaches for these and other emerging capabilities,” Bowman said.
Must ‘Strike Balance’
Bowman emphasized that regulators must strike a balance between ensuring safety and soundness and avoiding constraints that could hinder innovation. She said effective oversight of emerging technologies requires keeping pace with developments, coordinating across government agencies, communicating risks to supervised institutions, incorporating industry feedback, and continually refining supervisory responses.
“As we work to support innovation, it is necessary to determine whether our framework is appropriate,” Bowman said. “Have we established reasonable and effective supervisory expectations? Are bankers comfortable discussing emerging risks and new technologies with supervisory teams? Have we successfully implemented a pro-innovation mindset that allows responsible innovation and AI adoption to occur within the banking system?”
Report on Sound Practices is Coming
Bowman also said the Financial Stability Board’s Standing Committee on Supervisory and Regulatory Cooperation, which she chairs, is working with the U.S. Department of the Treasury and the Securities and Exchange Commission on a report outlining sound practices for AI adoption, use and innovation. A draft of that report is expected in the third quarter.
“The implications of AI extend far beyond the banking and financial systems,” Bowman said.
Separately, Anthropic said in April it had begun providing select partners with early access to a preview version of its Claude Mythos model, designed for defensive cybersecurity work, allowing organizations to identify vulnerabilities and strengthen systems before they can be exploited. The company also said it has been in discussions with government officials regarding the model’s offensive and defensive cyber capabilities.
