Federal Reserve Governor Michael Barr said artificial intelligence agents could set off a bank run in minutes, warning that automated financial decisions may move faster than banks can respond. Barr made the remarks in a speech at an AI event in Italy, according to prepared comments.
His concern centers on AI systems that can act on financial information without a human in the loop. If those systems start pulling deposits or selling assets in response to a signal, the resulting cascade could overwhelm a bank's ability to manage liquidity.
The speed problem
Traditional bank runs unfold over days or weeks, giving institutions time to reassure depositors or tap emergency funding. Barr's warning suggests AI agents could compress that timeline to minutes, leaving little room for the kind of human judgment that has historically slowed panics.
The issue isn't just speed. Barr also flagged that AI systems might bypass security controls, raising the prospect that automated tools could interact with banking infrastructure in ways regulators haven't anticipated. That combination — fast execution and weak guardrails — is what makes the scenario more than a thought experiment.
Why regulators are paying attention
Barr's comments land as financial regulators globally are trying to figure out how existing rules apply to AI. The Federal Reserve has been studying the technology's impact on banking stability, and Barr's speech adds a specific worry: that AI could turn a manageable disruption into a systemic event before supervisors even know what's happening.
Banks already use automation for trading, risk management, and customer service. But those systems typically operate within set parameters. AI agents that can make independent decisions — and learn from new data — introduce a different kind of risk, one that's harder to model and test.
What Barr didn't say
The governor didn't propose specific rules or timelines for addressing the risk. He didn't name any banks or AI developers. And he didn't offer a technical fix. Instead, his remarks read as an early warning, the kind of signal that often precedes formal guidance from the Fed or other regulators.
That leaves open the question of what comes next. The Fed could issue supervisory guidance, work with other agencies on standards, or simply continue monitoring. Barr's speech suggests the central bank is at least thinking hard about the problem.
The industry's response
Banks and AI developers haven't publicly responded to Barr's comments. The event in Italy was a speech, not a policy announcement, and there's no indication of immediate regulatory action. But the remarks are likely to be noted by compliance teams at large banks, who are already wrestling with how to govern AI tools.
For now, the warning stands as a marker. Barr has put the risk on the record. Whether it leads to new rules — or stays a cautionary note — depends on what regulators do next.




