Skip to main content
← All News

Markets · November 16, 2026 · 2 min read

Barr warns AI trading and fast deposit transfers could amplify financial stress

By Econ Data Tools Editorial Team

Published November 16, 2026 • Updated November 16, 2026

In November 16 remarks, the Fed governor described potential herding, cyber and consumer-finance risks. His financing figures are obligations and projections, not realized losses.

Federal Reserve Governor Michael S. Barr used a November 16, 2026 speech in Milan to examine potential financial-stability effects of artificial intelligence in securities trading, cybersecurity and consumer finance. Speaking at an Association for Computing Machinery conference on AI in finance, he described possible benefits alongside risks that could spread through markets. He said the views were his own, not necessarily those of the Board or the Federal Open Market Committee. Barr said generative AI could lower barriers to advanced trading technology and improve market surveillance. But widespread reliance on similar models and training data could synchronize trading strategies and reactions to news. He cautioned that shared flaws, manipulated information or concentrated market power could weaken market functioning. These are risk channels, not a finding that a specific AI-driven market collapse has occurred. He said his understanding was that generative AI was not yet being used for autonomous trading at large scale. Some theoretical research, he noted, finds that particular learning algorithms can develop collusive behavior or market-manipulation strategies. Those findings should not be generalized to every model or treated as evidence that all deployed trading systems act that way. In consumer finance, Barr raised concerns about conflicts in AI-generated investment advice, biased alternative credit data and the difficulty of explaining adverse lending decisions. He said major banks report AI agents attempting actions in retail accounts without banks being able to verify authenticity, consent and alignment with customer instructions. He called for strict authentication, action-specific consent checks and technical limits on agent capabilities. Barr also described a possible liquidity risk: AI agents could detect yield differences or signs of bank stress and move deposits quickly, potentially accelerating runs. That is a conditional warning about future or expanding use, not a report that an AI-triggered deposit run has already happened. On cybersecurity, he said AI can help both attackers and defenders, while increasing the importance of resilient shared infrastructure. The speech also recapped the latest Financial Stability Report. Barr said on-balance-sheet debt at seven large AI firms was expected to exceed $800 billion by year-end, and potential maximum nominal off-balance-sheet obligations exceeded $3 trillion. The latter include contracts, leases and guarantees and are not directly comparable to discounted on-balance-sheet debt or realized losses. He remained optimistic about AI's benefits but said institutions need legally precise, explainable and well-controlled use.

Sources

Related News