Which banks lean hardest on customers never bothering to move their money — and what happens when an AI agent does it for them?

This explores which banks depend most on customers leaving their money in place out of habit, and what happens to them if AI agents start moving cash automatically. One caveat first: the collection does not name or rank specific institutions, so it can't tell you which banks today are most exposed. What it does offer is more useful for thinking about the question. It shows that banks can depend on sticky deposits in two quite different ways, and agents threaten each one differently.

The first kind is dependence for profit. McKinsey's argument is that banks earn a lot from consumer passivity: savers leave money in low-yield accounts even when better rates are a click away, and cardholders stick with rewards programs they never compare Will AI agents break the deposit and credit card inertia?. These banks aren't fragile. Their business model simply assumes people won't bother to optimize. An agent that sweeps cash to the best rate removes that assumption. McKinsey estimates deposit margins could shrink by 30 to 50 basis points (0.30 to 0.50 percentage points) if only 10 to 20 percent of consumers hand the job to agents. The banks most exposed here are those with the widest gap between what they pay depositors and what the market pays.

The second kind is dependence for survival. Cullen Roche's point is that a bank with a weak balance sheet survives partly because withdrawals take days, which gives it time to react. Agents could compress those days into hours, so a run like the one that brought down Silicon Valley Bank (SVB) could happen faster Could AI agents speed up runs on individual weak banks?. Roche and Pompliano are clear about the limits of this. When agents move money between banks, total deposits in the system stay the same; the money just changes address. That makes agent-driven outflows a repricing problem for the system as a whole, not a collapse. Concentrated or weaker banks are the exception and face a narrower, sharper danger Would AI cash-sweeping agents trigger a systemic bank run?.

The surprising part is that these two groups may barely overlap. Large banks that profit from inertia stand to lose margin slowly. Smaller, concentrated banks that rely on slow withdrawals stand to lose depositors quickly. The same technology affects both, but one ends up with lower profits and the other with a liquidity crisis. There is also a real-world precedent for how fast frictionless saving can scale. China's Yu'e Bao, a money-market fund built into Alipay, reached 260 million users and $268 billion in assets by making it effortless to move idle cash into a higher-yield fund. McKinsey reads it as a warning about how fast agent-driven finance could spread, not as a technical breakthrough Could embedded financial optimization recreate Yu'e Bao at scale?.

One factor works against the idea that inertia will vanish quickly: people don't yet trust AI with their money. In a Harris Poll for Visa, 72 percent of US adults said they use AI assistants, but only 23 percent trust AI to make payments. Trust rose to 61 percent when a familiar brand like Visa was attached Does naming a payments brand increase trust in AI transactions?. That suggests stickiness will wear away gradually, and through trusted intermediaries rather than autonomous bots. If you want a list of specific banks, you'll need deposit-mix and rate data from outside this collection. What the corpus gives you is the two questions to ask about any bank: how much of its profit comes from depositor passivity, and how much of its survival depends on withdrawals being slow.