Banks Have Been Earning Effortlessly from Idle Funds for Years—Will AI Agents Change Everything?

According to Zhitong Finance APP, for many years, banks have profited from customers' idle funds, but AI agents may change this situation. Analysts have pointed out that Wall Street has a new concern about American banks: AI agents could help customers earn better returns on their cash.

Last week, Meta (META.US) released its new AI agent Muse, which quickly topped the US Apple App Store and immediately sparked a round of sell-offs in US bank stocks. The KBW Bank Index fell nearly 3% last Tuesday, with Charles Schwab (SCHW.US) tumbling 6% in a single day, leading the financial sector lower; JPMorgan Chase (JPM.US), Morgan Stanley (MS.US), and Wells Fargo (WFC.US) all dropped about 3%. Although bank stocks rebounded somewhat afterwards, investor anxiety did not dissipate.

The reason behind investor anxiety is that if AI agents enable consumers to effortlessly move idle money from 0.1% yield checking accounts to products yielding 5%—as easily as ordering takeout—banks’ moat of low-cost deposits will face unprecedented erosion. This is not a distant sci-fi story about technological substitution—it’s a risk for which pricing has already begun.

“Agentic Bank Run”: A warning that may sound sci-fi, but is mathematically real

Torsten Sløk, Chief Economist at Apollo Global Management, recently introduced a disturbing concept: the “agentic bank run.” In his report, he noted that AI assistants such as Muse may soon automatically transfer household cash from checking accounts earning an average of just 0.1% to fintech accounts offering interest rates between 3.3% and 5.0%.

Sløk outlined data showing that the FDIC national average savings rate is just 0.4%, with checking accounts even lower at 0.1%. Meanwhile, Adelfi offers a yield of 5.0%, SoFi (SOFI.US) 4.5%, LendingClub and LevelUp both 4.2%, and Pibank 4.1%. This is not a gap of a few dozen basis points—it’s a spread approaching five percentage points. For a $10,000 deposit, keeping it in a 0.1% checking account earns just $10 a year, while moving it to a 5% account brings in $500.

Although Sløk describes this as a “scenario that may soon occur” rather than a confirmed trend, if it becomes reality, the impact won’t be limited to a single bank. He wrote in his report: “If every household uses AI agents to optimize cash returns, banks may lose a large amount of the cheap deposits they rely on for lending. This would be a systemic issue for the entire financial system.”

The reason why this round of panic is so intense is because a significant portion of profits in US banking is built upon customer inertia.

Citrini Research provided a stunning figure in their September 22 report: Charles Schwab alone, by pricing customer cash and margin loans at lower rates than competitors, earns nearly $900 million in additional pre-tax profit each year. Specifically, Schwab pays just 0.19% on $246.35 billion in bank deposits, while Interactive Brokers (IBKR.US) offers a comparable rate of 2.23%. This difference alone provides Schwab with about $500 million in benefits annually. Adding the pricing differences on brokerage cash and margin loans, the total is about $899 million.

Citrini noted in the report: “Inertia rent exists because people are lazy.” Consumers tolerate low-yielding cash, useless subscriptions, and overpriced insurance simply because moving funds requires effort and attention—problems that AI agents simply do not have.

This is not just theoretical speculation. During the interest rate surge in 2023, Charles Schwab experienced a real “cash sorting” shock. From August 2022 to April 2023, clients moved funds en masse from low-yield accounts to high-yield money market funds, causing Schwab’s bank deposit balances to evaporate by nearly $50 billion, a 32% drop, forcing the company to tap expensive short-term financing and putting annual profits under pressure. The migration pace at that time was about $5.6 billion per month—but what an AI agent would do is essentially accelerate this process to just seconds.

The “wolf” hasn’t arrived yet, but the grass is already moving

It’s worth noting that not all Wall Street figures believe the end is near.

Bank of America analyst Ebrahim Poonawala stated in a client report on Thursday: “Until deposit costs rise at a pace exceeding what rates or competition can explain, the disruption narrative remains conceptual.” In other words, to prove that AI agents are truly changing customer behavior, there needs to be evidence of abnormal spikes in banks’ actual deposit costs—and that day has not yet come.

Executives at PNC (PNC.US) and Bank of America (BAC.US) have previously downplayed the threat of AI-driven cash optimization. Morgan Stanley analyst Manan Gosalia raised this topic directly with several bank executives at an industry conference in June, and the overall feedback was optimistic.

But the pressure of deposit competition does not come from AI—it was already heating up even before AI agents emerged.

The Federal Reserve has resumed rate hikes, which itself forces banks to offer depositors higher yields. Meanwhile, accelerated loan growth has made banks’ demand for deposits more urgent, while US personal savings rates hover around 2.6%, near a four-year low. These three forces are squeezing banks’ buffer space.

Although industry funding costs once fell from 2.61% in 2024 to 2.26% in 2025, this improvement is considered close to the end. Some analysts predict that if the Fed keeps rates high, deposit costs will enter a “plateau period”—and “the lack of rate cut prospects means deposit pricing pressure is real and mounting.”

Faced with a possible mass migration of deposits, big banks are not passively waiting.

The most noteworthy move comes from JPMorgan Chase. It is piloting an AI tool called Smart Cash, which can use AI models to predict clients’ cash flow needs, automatically shifting surplus funds from checking accounts into higher-yield brokerage products, while retaining enough daily liquidity. This tool first appeared in Jamie Dimon’s 2025 shareholder letter and was discussed further during the Q1 2026 earnings call. By June, the bank’s Chief Analytics Officer Derek Waldron had outlined plans to deploy AI agents capable of running independently for several hours.

From JPMorgan’s perspective, if customers’ money is destined to be moved by AI, it might as well be moved by the bank’s own AI—at least the money stays within their system. The bank has already recategorized about $2 billion in AI-related expenditures as core infrastructure, with AI representing about 10% of its $19.9 billion tech budget for 2026.

Other major banks are also strengthening their defenses. Last week, Citi launched a new savings rate program called “Citi Premium Boost,” targeting Citi Priority, Citigold, and Citigold Private Client tiers, which unlocks higher savings rates based on client status and qualifying banking activities; new customers are automatically enrolled from October 26. Previously, PNC and Bank of America also introduced similar incentives to attract higher client balances.