Apollo Economist Warns AI Agents Could Trigger Slow-Motion Bank Run
AI agents may shift household cash away from banks
Apollo Global Management Chief Economist Torsten Slok has warned that artificial intelligence agents could trigger a "slow-motion bank run." He argues these automated systems might move household cash from low-interest checking accounts to higher-yield alternatives.
In a note published on Sunday, Slok suggested that widespread adoption of these agents could drain the inexpensive deposits banks rely on to fund loans. This shift could pose a threat to the broader financial system.
How agentic finance changes deposit dynamics
- Current checking accounts offer a national average interest rate of just 0.1%.
- Firms like Revolut, SoFi, and Wealthfront pay between 3.3% and 5% on deposits.
- A $10,000 balance earns roughly $100 annually at 5%, compared to only $10 in a standard checking account.
- Agentic AI refers to software that acts autonomously, monitoring balances and moving funds in real time to optimize returns.
The warning from Apollo's chief economist
Torsten Slok, a partner at Apollo which manages about $1 trillion in assets, highlighted the disparity in interest rates. He noted that Meta's personal agent, Muse, and similar tools could soon automate the transfer of idle cash.
Slok stated, "If every household used AI agents to optimize the return on their cash balances, banks could lose a large share of the cheap deposits they rely on to make loans, which would be a problem for the entire financial system."
Crypto infrastructure supports automated payments
Crypto payment networks are already developing the infrastructure required for these autonomous agents. Coinbase's x402 protocol allows AI agents to pay for online services using stablecoins—digital currencies pegged to fiat values—in seconds without human approval or traditional banking cards.
Reports indicate the x402 Protocol has processed between 188 million and 205 million cumulative transactions to date.
Market growth estimates vary
Projections for the size of the agentic AI market in financial services differ significantly. Mordor Intelligence estimates the sector will reach $7.78 billion in 2026, growing to $43.52 billion by 2031. In contrast, MarketsandMarkets sizes the narrower AI agents segment at approximately $845 million in 2025.
Why this matters for the financial system
The concern centers on how banks generate credit. They typically use low-cost customer deposits to fund loans. If AI agents rapidly move these funds to higher-yield accounts, banks may face difficulties in maintaining the liquidity needed for lending, potentially destabilizing the system.
Where to read more
Newisty Editorial Team covers technology, cryptocurrency, digital products, online platforms, developer tools and the wider digital economy. Our content is researched from official sources, company announcements, public documentation, market data and other primary or reputable sources. Articles are reviewed and edited before publication for clarity, accuracy and useful context.