Apollo warns of 'agentic bank runs' as AI labs admit they can't stop agents from acting in unison

Apollo Global Management chief economist Torsten Slok has warned about a threat to the banking system that he refers to as an “agentic bank run.”

The concern is that if millions of people permit AI agents to move their savings to chase better interest rates, nothing stops all those agents from acting in the same week.

How are AI agents a threat to banks?

In a note published on September 27, the chief economist at Apollo Global Management, Torsten Slok argued that consumer AI agents could soon take all the household cash out of low-yield checking accounts and put them into fintech accounts that pay far more.

A $10,000 checking balance at the 0.1% national average earns roughly $10 a year, but the same money in a 5% account would earn $500. Using data from the FDIC, Haver Analytics and providers including Revolut, Varo Bank and Wealthfront, Slok listed 11 fintech and online accounts paying between 3.3% and 5%, with Adelfi at the top at 5% and SoFi close behind at 4.5%.

Banks keep checking rates near zero because those deposits are their cheapest funding for loans, and retail banking assumes most of that money stays put.

Slok warns that banks will lose the cheap base that makes lending profitable if AI agents begin moving substantial amounts.

The Bank of England also warned in its April 2025 Financial Stability in Focus paper that advanced AI trading strategies could push firms into “increasingly correlated positions and acting in a similar way during a stress, thereby amplifying shocks.”

The paper defines agentic AI as systems that take autonomous action toward set goals by using tools and adapting to their environment.

What AI agents can make bank transfers?

The trigger Slok’s note points to is Meta’s Muse, which was launched on September 8 with account connections powered by Plaid.

Muse can currently view a user’s balances, transactions and investment holdings, but it cannot yet initiate transfers.

Muse was shipped in the same week that Jacob Coxon, a researcher who had worked at both OpenAI and Anthropic, quit the field and accused both companies of “gambling with our lives.”

Cryptopolitan previously reported that Meta’s leadership went on to push back against calls for an industrywide slowdown. In contrast, at the UN Security Council on September 23, Anthropic’s Dario Amodei pledged to “slow down as much as necessary” and OpenAI’s Sam Altman called for “extreme care.”

The idea that AI could synchronize market behavior was first raised in a 2023 speech when the then-SEC Chair Gary Gensler warned that a few dominant AI base models could make “similar decisions because they are getting the same signal.”

Notably, the Federal Reserve raised its target range to 3.75% to 4% on September 16, its first hike in three years, and bank stocks fell afterwards due to investor concerns that a flatter difference between short- and long-term rates would squeeze how much profit banks make on loans.

If you're reading this, you’re already ahead. Stay there with our newsletter.