The paperExchangeCrypto WikiWikiAcademynotitiaplatform.com
Sign in
● LIVE
CAC 408 281,4▼ 0.85%S&P 5007 416,6▼ 1.92%EURO STOXX6 289,7▼ 0.92%BTC/USD64 206▼ 0.44%ETH/USD1 792,7▼ 0.69%EUR/USD1.10▼ 0.03%BRENT92.45▲ 0.01%GOLD3 861,4▲ 0.22%OAT 10Y5.12%▼ 0.64%VIX23.56▲ 2.84%CAC 408 281,4▼ 0.85%S&P 5007 416,6▼ 1.92%EURO STOXX6 289,7▼ 0.92%BTC/USD64 206▼ 0.44%ETH/USD1 792,7▼ 0.69%EUR/USD1.10▼ 0.03%BRENT92.45▲ 0.01%GOLD3 861,4▲ 0.22%OAT 10Y5.12%▼ 0.64%VIX23.56▲ 2.84%
 
Europe edition
Notitia
← Back
TechnologyCryptoast

Fed warns: AI agents could speed up future bank runs

Governor Michael Barr of the US Federal Reserve (Fed) is concerned about the consequences of advances in artificial intelligence (AI). The central banker believes that autonomous AI agents could accelerate future bank…

C
Cryptoast
Newswire · 20 min ago
◌ Single source2 min read
✦ Explain it to me
Fed warns: AI agents could speed up future bank runs

Home » Crypto News » Economy and Finance News » The Fed issues a warning: AI agents could accelerate future bank runs

https://cryptoast.fr/fed-michael-barr-agents-ia-accelerer-bank-run/

During an international conference on artificial intelligence held in Milan (Italy) on November 16, 2026, Michael Barr, member of the Board of Governors of the Federal Reserve, warned that autonomous AI agents could dangerously accelerate bank runs, that is to say massive and sudden bank withdrawals.

In his speech, Governor Barr explains that these AI tools would be able to instantly detect yield variations or stability alerts. Thus, where deposit outflows historically required manual action by each client, autonomous agents could accelerate these outflows to the point that they “happen in minutes, outpacing traditional liquidity management and regulatory intervention timelines.”

Speech by Governor Barr on some potential financial stability implications of artificial intelligence: https://t.co/5WJ28iremn

In other words, a bank panic with AI-driven acceleration could empty a bank before the regulator even has time to react. This, while the speed of digital transactions has already complicated the task. We will particularly remember the case of the fall of Silicon Valley Bank (SVB), on March 9, 2023: customers attempted to withdraw $42 billion in a single day, or nearly a quarter of the establishment's total deposits, precipitating its placement under supervision the next day.

This bank run, already considered the fastest in modern American banking history, nevertheless took place over several hours via manual orders placed from mobile applications. Autonomous AI agents could compress this same movement in a few minutes, well before any intervention from the regulator.

Faced with the risk of massive and sudden withdrawals, taking stock of the security, transmission and taxation of one's assets can prove useful.

At least five major institutions, including JPMorgan Chase, Citigroup, Bank of America, Wells Fargo and Morgan Stanley, now block third-party AI agents from accessing their clients' accounts.

On the same subject – AI agents: major American banks are trying to block unauthorized access

Beyond bank run risk, Michael Barr points to broader systemic vulnerabilities linked to the AI ​​infrastructure boom. According to the Fed's latest Financial Stability Report, debt on the balance sheets of the seven largest AI companies is expected to exceed $800 billion by the end of the year. Off-balance sheet commitments, including long-term data center leases and energy supply contracts, now exceed $3 trillion.

The governor also mentions a phenomenon of “circular financing” where large chip manufacturers (like Nvidia) and hyperscalers (cloud giants, like Amazon and Microsoft) inject more than $100 billion into small start-ups, which reuse these funds to purchase computing capacity. A pattern that could amplify a market shock if returns on investment in AI were to disappoint.

Michael Barr also cites asymmetric risks in terms of cybersecurity. Autonomous models would tend to treat safety barriers as “obstacles to overcome rather than rules to respect”. A problem already observed in the crypto ecosystem, where AI agents have managed to circumvent their own spending limits.

On the same topic – AI agents bypassed their crypto spending limits

The warning messages from the Fed Governor can thus be summed up by the fact that the adoption of agentic AI in finance is moving faster than the safeguards, and regulators must catch up before the next banking shock reveals the full extent of the problem.

To track the companies at the heart of the Fed's AI infrastructure boom, here's a chart of AI stocks around the world.

Already intrigued by Bitcoin and blockchain technology in 2013, I professionalized myself in the sector by becoming, since 2018, an editor specializing in crypto news. I followed the cryptosphere through its cycles, from early amateur mining to the gradual structuring of the sector. Now independent, I cover crypto news, financial markets and regulation.

The site that explains everything from A to Z about Bitcoin, blockchain and cryptocurrencies. News and explanatory articles to discover and progress in these sectors!

Nvidia: Billionaire Leo KoGuan will no longer invest without Jensen Huang

Binance France will close your account: 4 things to do before November 29

Bitcoin Crashed to $62,000 Overnight – What Happened?

Following your registration and depending on the amount of the deposit

CFDs are complex instruments and carry a high risk of rapid loss of capital due to leverage. 55% of retail investor accounts lose money when trading CFDs with this provider.

Register with the code CRYPTOAST and trade at least €1,000 before 10/31

8% per year on your crypto deposits for 3 months - Visit

For new and old customers residing in France, after identity verification

Source · Cryptoast ↗

Machine-translated from French

Not cross-checked yet

Only Cryptoast reports this story so far. We show it as a single source until another newsroom confirms it.

The reports
  1. Cryptoast
    17 Nov, 11:30
Trust score: 74/100
Publisher Cryptoast (general tier)+72
Detailed summary+2
Total74

Method: reports are grouped by the words of their headline and summary; each newsroom counts once; figures are read from the reports as published. Nothing here is written by an AI.