Federal Reserve Governor Michael S. Barr has said that AI agents could move funds so quickly in a stress period that bank runs might unfold in minutes, overwhelming the timelines on which banks and regulators now manage liquidity.
In a 16 November 2026 speech at the Seventh Annual Association for Computing Machinery International Conference on Artificial Intelligence in Finance, in Milan, Barr set out what he called “some potential financial stability implications of artificial intelligence.” A footnote to the speech states: “The views expressed here are my own and are not necessarily those of my colleagues on the Federal Reserve Board or the Federal Open Market Committee.”
Where customers once needed to execute transfers themselves, Barr said, “AI agents could instantaneously detect yield differentials or stability signals and autonomously move funds at high speed, potentially accelerating runs that unfold in minutes, overwhelming traditional liquidity management and regulatory intervention timelines.”
What Barr said banks are already seeing
The run-speed warning sat in a wider discussion of agentic systems in finance. Barr said new risks continue to surface. “For example, major banks now report finding AI agents attempting to take actions in retail customer accounts, without the bank being able to verify authenticity, consent, and alignment with customer instructions.”
Those incidents, he said, “underscore that AI in financial services can only succeed long term with appropriate guardrails: strict authentication, explicit consent verification for specific actions, technical controls limiting agent capabilities, and both rules- and values-based constraints.” He added that technological innovations are needed to ensure AI agents follow established rules and values, “especially critical as agentic commerce expands.”
Barr also warned about herding in markets. If traders rely on “a small set of similar AI models trained on similar data,” he said, they face the risk of herding behaviour, in which trading strategies become similar and reactions to news synchronise. Algorithm flaws can cause failures, he said, but a market with large numbers of similar models that may share the same flaw risks failures with systemic consequences. Bad actors, he added, could intentionally manipulate trades, quotes or news and potentially mislead similar AI trading systems at the same time.
On cyber risk, Barr said incidents this year had shown risks from AI models acting autonomously, not in alignment with the goals of their human creators. “Given a goal, AI models have treated security boundaries as obstacles to overcome rather than rules to be followed, raising questions about how effective rules will be even against AI agents instructed to follow them.”
The speech was not only a catalogue of hazards. Barr began by calling himself “an optimist about the long-term promise of AI to empower individuals, businesses, and institutions across society,” and said he remained hopeful that, if today’s challenges are managed, AI will raise living standards. He closed by arguing that opportunities abound for using AI to improve access and efficiency, provided that this “proceeds in ways that protect users and support financial stability.”
Barr told the conference that there are “important insights about AI and financial stability in the Federal Reserve’s latest Financial Stability Report (FSR) issued on Friday.” That November 2026 report, he said, focuses on the AI investment boom and on staff research into leverage in some of the debt structures used to finance that growth. He then turned to his own discussion of how AI is being integrated into financial markets. The FSR is a Board publication; the agentic-run remarks that followed are his.
The speech sits alongside other official financial-stability work TSN has covered, including the Federal Reserve’s latest Financial Stability Report, whose survey contacts named AI as the top-cited risk, the Financial Stability Board’s sound practices for responsible AI, and the IMF’s October 2026 Global Financial Stability Report, which listed an AI market correction among near-term risks. Those documents are separate from Barr’s Milan speech.
A Milan speech is not a Federal Reserve policy
Barr’s remarks are a speech by one governor, delivered at an academic conference, and caveated in his own footnote as his views rather than those of the Board or the FOMC. They are not a supervisory guidance document, not a regulation, and not a vote. Reading them as a change in Federal Reserve policy would go beyond what he published.
The “runs that unfold in minutes” passage is itself conditional. Barr described a mechanism — agents that can detect yield or stability signals and move funds without a customer executing the transfer — and said this could accelerate outflows and overwhelm existing timelines. He did not present a historical case study of such a run, a quantitative model of how short those minutes might be, or a proposed new liquidity rule. The gap between a described mechanism and a measured event is left open because he did not fill it.
The retail-account incidents are similarly bounded. Barr said major banks now report finding agents attempting to take actions without the bank being able to verify authenticity, consent and alignment with customer instructions. He did not name the banks, count the incidents, or say those attempts succeeded in moving funds. The guardrails he listed are a prescription for how, in his view, agentic finance can succeed over the long term, not a checklist the Board has adopted.
Herding and the “security boundaries as obstacles” remarks are of the same kind: reasoned risks, with citations in the speech’s footnotes, offered as reasons for caution. They should be read as one governor’s account of those risks, not as an official finding that markets have already stampeded or that existing rules have already failed.
Sources
- Michael S. Barr, “Some Potential Financial Stability Implications of Artificial Intelligence,” speech at the Seventh Annual Association for Computing Machinery International Conference on Artificial Intelligence in Finance, Milan, 16 November 2026
- Board of Governors of the Federal Reserve System, Financial Stability Report, November 2026
Get TSN’s biggest tech stories, explained.
News and insights across AI, tech, crypto and cybersecurity. Every source linked. Unsubscribe anytime.
Get the newsletterRelated stories
- New York Fed Survey Contacts Name AI as the Top-Cited Stability Risk
- FSB Finalises 12 Sound Practices for Responsible AI
- IMF Names an AI Market Correction, Rising Bond Yields and Cyberattacks as Near-Term Risks to Financial Stability
Follow @tsnmediahub on X for breaking AI news as it happens.
Follow on X