AI agents could drain cheap bank deposits, Apollo's Torsten Slok warns
Reported by CoinDesk · AI-assisted summary by ChikoCorp AI News Desk
AI-generated summary based on the linked source; not independently verified. This is not investment advice. Verify market-moving details at the original publisher before acting. See our editorial policy, AI content policy, and financial disclaimer.
Summary
Apollo Global's Chief Economist Torsten Slok warned that AI-driven financial assistants could cause a slow-motion bank run by automatically shifting household cash from low-interest checking accounts to higher-yield alternatives. This shift could drain the inexpensive deposits banks use to fund loans, potentially threatening the broader financial system. Slok pointed to AI agents like Meta's Muse and fintech firms paying interest rates between 3.3% and 5% on deposits as drivers of this change.
Why it matters
The development matters because widespread use of agentic AI assistants could disrupt traditional banking by reducing banks’ access to cheap deposits, which are critical for lending. This could pose risks to the financial system as banks lose low-cost funding sources. The source highlights that this impact arises from AI agents' ability to optimize household cash management in real time.
Key context
Agentic finance refers to AI systems that act autonomously, managing funds by monitoring balances, comparing returns, and moving money between accounts without human intervention. The market for agentic AI in financial services is projected to grow significantly, with estimates ranging from $845 million in 2025 to nearly $44 billion by 2031. Additionally, crypto payment rail infrastructure like Coinbase's widely used x402 protocol is enabling instant payments by AI agents, supporting this agentic finance ecosystem.
Key numbers and entities
Torsten Slok is Apollo Global’s Chief Economist and partner; Apollo manages about $1 trillion in assets. Interest rates from fintech firms such as Revolut, SoFi, Varo, LendingClub, and Wealthfront range from 3.3% to 5%, compared to a 0.1% average on checking accounts. The x402 protocol has processed between 188 million and 205 million transactions, with about 69,000 active AI agents. The x402 Foundation includes Cloudflare, Google, Visa, Mastercard, AWS, Circle, and Stripe, under Linux Foundation governance.
What remains unclear
The source does not specify the timeline over which AI agents might shift household cash and materially affect banks' deposit bases. It also does not discuss regulatory responses or whether banks might adapt to this shift. Further details on how households would deploy agentic AI tools and the exact scale of impact on the financial system are not provided.