Federal Reserve Governor Barr Addresses AI Challenges for Banks
Federal Reserve Governor Michael S. Barr raised critical questions for banks eager to adopt AI agents, particularly regarding how to ensure these agents fulfill customer requests accurately. He pointed out that banks will encounter significant challenges in confirming that agentic AI adheres to customer instructions, a concern that resonates deeply within an industry reliant on authorization and liability.
For instance, when a customer instructs an AI agent to "pay my bills," the agent must determine which bills to pay, from which account, in what order, and by what deadline. Each decision point presents an opportunity for misalignment between customer intent and the agent's actions.
Market Reading & Resources
Sponsor funding complete. Trade on.
During his address at the Singapore Fintech Festival on November 11, 2025, Barr noted that major banks are exploring the use of generative AI for essential functions such as credit assessments, fraud detection, and trading. However, he set a high standard for these technologies, insisting that their outputs must be controlled, precise, explainable, and replicable. Failure to meet these criteria poses significant risks, including potential biases in lending and unintended consequences like increased market volatility from trading algorithms.
In light of these concerns, four major U.S. banks—JPMorgan Chase, Citigroup, Charles Schwab, and Wells Fargo—announced between October 13 and 20, 2026, that they would restrict third-party AI agents from accessing customer accounts. The banks cited worries over unauthorized transactions and unclear liability for actions taken by these so-called rogue agents. Notably, Wells Fargo has updated its terms to shift liability for unauthorized actions by AI agents onto customers, effective November 24, 2026.
Banks Prioritize Internal AI Tools Amid Regulatory Guidance
As of late 2026, many banks are developing their own internal agentic tools while limiting the use of external agents. In April 2026, U.S. interagency guidance explicitly excluded generative and agentic AI from model risk management requirements, leaving banks to manage these risks through internal governance and contractual adjustments.
Regulatory Scrutiny on AI Accountability in Banking
Barr's remarks signal a shift in supervisory focus, suggesting that regulators may soon scrutinize how banks ensure that AI agents operate within the bounds of customer intent. For fintech companies developing consumer AI agents, the recent restrictions from major financial institutions represent a significant barrier to accessing customer accounts.
Ultimately, the implications for consumers revolve around accountability. The changes in Wells Fargo's terms indicate that, in the absence of regulation, customers may bear the financial consequences of errors made by AI agents. Observers will be watching closely to see if the Federal Reserve or other regulatory bodies take steps to incorporate agentic AI into formal risk management frameworks and whether more banks will follow Wells Fargo's lead in transferring liability to customers.




