Fed Vice Chair Bowman Outlines FSB Sound Practices for Responsible AI Adoption in Financial Sector
Key Takeaways
- •The FSB consultation report on responsible AI adoption in financial services is expected to be finalized later in 2026 as a U.S. G-20 deliverable after a public comment period.
- •Bowman identified materiality and proportionality as core principles, stating that lower-risk AI uses should receive lighter regulatory scrutiny while governance should scale with an institution's size and complexity.
- •The Federal Reserve has been monitoring AI usage by banks for nearly a decade and has observed growing adoption across institutions of all sizes and across diverse use cases.
- •Bowman encouraged stakeholders to provide feedback on whether the report is too prescriptive, insufficiently tailored, or missing material risks that warrant additional attention.
- •The report was produced through international collaboration led by Hern Shin Ho of the Monetary Authority of Singapore, with input from U.S. colleagues at the Treasury Department and the SEC.
Federal Reserve Vice Chair for Supervision Michelle W. Bowman delivered opening remarks on July 7, 2026, at the Financial Stability Board (FSB) Virtual Outreach Event, discussing the FSB's consultation report on Sound Practices for Responsible Adoption of Artificial Intelligence. The FSB, which coordinates international financial regulation across G-20 jurisdictions, developed the report as part of a broader global effort to address the systemic implications of AI adoption in financial services. The report is expected to be finalized later this year as a U.S. G-20 deliverable, and both Federal Reserve staff and the FSB Secretariat are seeking public feedback to inform the final version.
Bowman, who chairs the FSB's Standing Committee on Supervisory and Regulatory Cooperation, initiated work on the sound practices framework late last year as part of her committee priorities. She noted that since then, AI has continued to evolve rapidly and financial institutions' use of the technology has expanded, making public engagement on both benefits and risks increasingly important.
Bowman credited Hern Shin Ho of the Monetary Authority of Singapore for leading the FSB work stream that produced the report within a compressed timeframe. She also expressed gratitude to FSB Secretariat staff and to U.S. colleagues at the Treasury Department and the Securities and Exchange Commission (SEC) for their collaboration.
The Federal Reserve has been monitoring bank usage of AI for nearly a decade, Bowman said, observing a noticeable increase in adoption across banks of all sizes and across a wide range of use cases. This trend parallels similar scrutiny from other U.S. banking agencies, including the OCC and FDIC, which have issued interagency guidance on AI-related risks such as model risk management. She emphasized that the Fed's focus has been on supporting institutions seeking to innovate responsibly by leveraging AI tools, and that domestic supervisory experience helped inform the FSB report.
A central element of managing AI risks, according to Bowman, is understanding specific use cases. The FSB report includes a variety of examples and in-depth case studies illustrating types of governance and controls that may be appropriate in different scenarios. However, she stressed that these practices represent one approach, not the only path to responsible AI adoption and risk management.
Bowman highlighted materiality as a key consideration, stating that financial institutions should be specific about how they deploy AI and whether such use is material to their business operations or legal and regulatory obligations. The actual use and its materiality, she said, should inform the type and intensity of governance and controls applied. The report emphasizes that lower-risk uses of AI warrant a lighter supervisory and regulatory touch.
Proportionality is another significant focus of the report. Bowman noted that approaches suitable for larger institutions deploying AI in complex applications may not be appropriate for smaller institutions with simpler AI uses. She reiterated that the Fed's objective is to promote innovation at financial institutions of all sizes, not only the largest ones, and the report is designed to provide clear guidance to institutions across the spectrum. She invited feedback on whether the report achieves the right balance on proportionality.
Bowman described the report as an important first step and encouraged respondents to identify areas where the sound practices may be too prescriptive or insufficiently tailored to differences in institutional size, complexity, and risk profile. She also asked stakeholders to point out where material risks may not have been adequately addressed or where additional clarity would help institutions manage AI-related risks more effectively.
The feedback gathered during the event and throughout the public comment period will inform the final report, which will be delivered to the U.S. G-20 presidency later this year. The report's delivery as a G-20 deliverable positions it to inform ongoing international policy discussions on AI governance in the financial sector. Bowman expressed satisfaction with the international collaboration on the initiative and indicated she looks forward to further improvements in the final report.
Bowman noted that the views expressed in her remarks are her own and not necessarily those of her colleagues on the Federal Reserve Board or the Federal Open Market Committee.