The views expressed in these remarks are those of the speaker in his role as Chairman of the FSB Standing Committee on Supervisory and Regulatory Cooperation. and do not necessarily reflect the views of the FSB or its members.
Thank you for the opportunity today to discuss the Financial Stability Board (FSB) Consultative Report on Sound Practices for the Responsible Implementation of Artificial Intelligence. Fed staff and FSB Secretariat staff are keen to consider and incorporate feedback on the consultation report so that it can be finalized as a U.S. G20 product later this year.
Before I offer some thoughts on the report itself, I would like to thank Mr Hearn Singh Ho from the Monetary Authority of Singapore. Hern Shin led the FSB’s work flow that produced this report in a very short period of time. I would also like to pay tribute to and thank the staff of the FSB Secretariat who supported this work. Finally, I would like to thank my U.S. colleagues at the Treasury Department and the SEC for their assistance in preparing this report.
As many of you may know, I am the chair of the FSB’s Standing Committee on Supervisory and Regulatory Cooperation. When setting priorities late last year, I began working towards a set of sound practices that financial institutions should consider when responsibly adopting and using artificial intelligence (AI). Since then, AI has continued to evolve rapidly, and financial institutions’ use of AI has expanded. Given these ongoing developments, it is important that our report allows the public to discuss the potential benefits of AI, as well as the risks and other challenges that may arise from its use. The report also describes how financial institutions, including banks, are already successfully addressing and managing AI use risks, with a broader goal of responsible innovation that benefits financial institutions and their customers.
The Fed has been monitoring banks’ use of AI for nearly a decade. There is a noticeable increase in the use of AI in banks of all sizes, and we are seeing a wide range of use cases. We are focused on supporting institutions that want to innovate responsibly by leveraging AI tools in their operations. Our work in the United States helped inform the FSB report.
A central element of AI risk management is understanding your specific use case. The FSB report provided a range of examples, including several detailed case studies, to illustrate the types of governance and controls that would be appropriate in similar circumstances. To be clear, these practices are not the only way to deploy and use AI responsibly or manage AI risk.
Another important aspect to highlight in this report is whether the use of AI matters. From my perspective, financial institutions need to be specific about how they use AI and whether it is material to their business operations or legal and regulatory obligations. Its actual usage and importance can help inform the type and strength of governance and controls that should be applied to a particular AI deployment. In this report, we highlighted that low-risk uses of AI should be subject to less oversight and regulation.
Similarly, the report also focuses on proportionality. What works or considerations for large institutions using AI in complex applications are not appropriate for smaller institutions with less complex uses of AI. As I mentioned earlier, we are focused on driving innovation across financial institutions of all sizes, not just the largest. This report provides clear guidance for all institutions, including small institutions. We therefore welcome feedback on whether this report strikes the right balance in terms of proportionality in relation to the adoption and use of AI.
Finally, this report is an important first step in considering how financial institutions can use AI responsibly. When providing feedback, we encourage you to identify areas where sound practices may be too prescriptive or fail to adequately account for differences in organization size, complexity, and risk profile. Equally important, we ask you to help us identify where we may not be adequately addressing significant risks or where further clarification could help institutions more effectively manage AI-related risks. We need to ensure that these practices support responsible innovation across the financial system, while maintaining appropriate safeguards, especially for high-risk applications.
The feedback we receive today and during the public comment period will be reflected in the final report submitted to the U.S. G-20 Presidency later this year. I am pleased with the international collaboration so far on this research and look forward to improving this in the final report later this year.
