Skip to main content
Live
Main content

Fed Governor Barr warns AI agents could turn bank runs into minutes-long events

Speaking in Milan, Barr said banks already see AI agents acting in customer accounts without verified consent, and called for strict authentication.

Jaeden Schafer
Editor in Chief · · 3 min read
Fed Governor Barr warns AI agents could turn bank runs into minutes-long events

Federal Reserve Governor Michael Barr warned on Monday that autonomous AI agents could accelerate bank runs to the point where they unfold in minutes. Barr made the case in a speech titled "Some Potential Financial Stability Implications of Artificial Intelligence," delivered at the Seventh Annual Association for Computing Machinery International Conference on Artificial Intelligence in Finance in Milan, Italy.

“AI agents may accelerate deposit outflows during stress periods.”
— Michael S. Barr, Federal Reserve Governor

The core concern is speed. Barr said AI agents could instantaneously detect yield differentials or stability signals and autonomously move funds at high speed, overwhelming the liquidity management and regulatory intervention timelines that banks and supervisors rely on today.

Barr said the issue is already showing up in practice. According to the speech, 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 the customer's instructions.

Key facts

  • 01Fed Governor Michael Barr said on Nov. 16 that AI agents could accelerate bank runs that unfold in minutes.
  • 02Barr said major banks report AI agents attempting actions in retail accounts without verified customer consent.
  • 03He called for strict authentication, action-specific consent checks and technical limits on agent capabilities.
  • 04Barr said on-balance-sheet debt of the seven largest AI firms is expected to exceed $800 billion by year-end.
  • 05Barr put off-balance-sheet liabilities at more than $3 trillion in the same speech.

His proposed answer is a set of controls rather than a ban. Barr called for strict authentication, explicit consent verification for specific actions, and technical controls limiting what agents are able to do inside financial accounts. He did not announce a new Fed rule or supervisory guidance.

The speech also covered the financing behind the AI buildout. Barr said on-balance-sheet debt of the seven largest AI firms is expected to exceed $800 billion before year's end, and put off-balance-sheet liabilities at more than $3 trillion. Hyperscalers and GPU makers hold over $100 billion in cumulative equity investments in large language model labs and cloud start-ups, he said.

Barr tied those figures to the Federal Reserve's Financial Stability Report, which the central bank issued on Friday.

On cybersecurity, Barr pointed to cases where AI models have treated security boundaries as obstacles to overcome rather than rules to be followed.

Barr also set a bar for how financial institutions use AI internally. Decisions based on AI models, he said, must be well controlled, numerically and legally precise, explainable, and replicable.

“I believe we can and must ensure that this proceeds in ways that protect users and support financial stability.”
— Michael S. Barr, Federal Reserve Governor
Related · from this week
DeepSeek says its Harness agent app hit 1.54M daily users
Jaeden Schafer · 3 min read →

Barr framed the guardrails as a condition for AI adoption in finance rather than an argument against it, saying AI in financial services can only succeed long term with appropriate guardrails.

ShareXLinkedInEmail
AI Box

Every AI model. One chat.

The latest models from ChatGPT, Claude, Gemini, Sora, ElevenLabs — 80+ models in a single chat. Compare answers side by side. Pick the best one every time.

  • ChatGPT, Claude, Gemini, Grok, DeepSeek — in one chat
  • Generate images & video with Sora, Veo, Ideogram
  • Compare any two models side by side
  • From $8.99/mo · 80+ models, all included
Try AI Boxaibox.ai
Trusted by 3,000+ teams
Got a tip?

Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.

Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.

AI Box Daily briefingFree · Daily · No fluff

Stay ahead of everyone in AI.

The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.

Loved by 10,000+ AI professionals
Free forever. Unsubscribe with one click.

The briefing read inside teams at

Keep reading

More from News

DeepSeek says its Harness agent app hit 1.54M daily users
News

DeepSeek says its Harness agent app hit 1.54M daily users

Daily users grew from about 271,000 at the end of September, and DeepSeek is now preparing an official plugin marketplace for the app.

Jaeden Schafer3 min read
Anthropic logo
News

Anthropic study finds analysts follow their AI agent's lean 86% of the time

In a 50-analyst experiment, a coin flip set which way Claude leaned on each call, and most experts went along without crediting the agent.

Jaeden Schafer5 min read
Meta logo
News

Meta launches Muse, a personal AI agent built on a security-first pitch

The Superintelligence Labs product ships on iOS, Android, and WhatsApp with Stripe checkout and a $300,000 bug bounty ceiling.

Jaeden Schafer5 min read