Apollo Chief Economist Torsten Slok warned Sunday that AI assistants managing household cash could trigger a slow-motion bank run by sweeping deposits from low-interest checking accounts into higher-yield products.

In a note titled "Is an Agentic bank run coming?" Slok argued that Meta's Muse and comparable agentic assistants could reroute balances automatically, bypassing the 0.1% national average on checking accounts.

"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," Slok wrote.

How big is the deposit gap?

Slok, a partner at Apollo Global, which manages roughly $1 trillion in assets, highlighted a stark rate differential between incumbents and challengers. Revolut, SoFi, Varo, LendingClub and Wealthfront pay between 3.3% and 5% annual interest on deposits. On a $10,000 balance, that produces roughly $100 a year in interest, versus about $10 from a typical checking account.

That gap is the mechanism through which agentic AI could become a structural threat rather than a consumer convenience. Banks fund loans primarily with inexpensive deposits; an automated, large-scale migration of retail cash would compress that funding base.

What does "agentic" actually mean here?

Agentic finance refers to AI systems that act rather than merely respond. Such agents monitor balances in real time, compare returns across institutions, sweep idle cash into higher-yield accounts and reposition funds ahead of scheduled bills, all without per-transaction human approval.

Slok's warning frames a near-term deployment risk rather than a speculative one. Meta's consumer agent is already in market; competing products from incumbents and fintechs are following.

What is the market opportunity?

Estimates diverge sharply. Mordor Intelligence sizes agentic AI in financial services at $7.78 billion in 2026 and projects $43.52 billion by 2031. MarketsandMarkets puts the narrower AI agents segment at approximately $845 million in 2025.

The disparity reflects differing taxonomies, but both confirm that capital is flowing into the category. Slok's concern is that adoption, not innovation, is the variable that matters for banks.

Where does crypto fit?

Crypto infrastructure is positioning itself as the payment rail for agentic finance. Coinbase's x402 protocol, the most widely used agentic payment standard, lets an AI agent settle for online services in stablecoins within seconds, without an account, card or human sign-off.

Cumulative x402 transactions have reportedly reached between 188 million and 205 million, with roughly 69,000 active agents. Cloudflare, Google, Visa, Mastercard, AWS, Circle and Stripe have joined the x402 Foundation, which the Linux Foundation now governs.

ETF Institute co-founder Nate Geraci has previously argued that AI and crypto are both converging on the traditional banking model from different angles.

What changes for banks?

The operational risk is twofold. First, deposit attrition would push banks toward wholesale funding, raising the cost of credit. Second, automated, condition-based withdrawals could accelerate outflows during stress, compressing the time banks have to react.

Slok's note does not propose a specific policy response, but it implicitly elevates deposit-fragility monitoring to a front-line concern for risk committees. Regulators have so far focused on stablecoin reserves and custody; agentic deposit migration sits outside that perimeter.

The forward question is whether retail deposit insurance, disclosure standards, and rate-negotiation guardrails will be drafted to govern autonomous cash movement before adoption reaches the scale Slok describes.