Originally written as an OpEd, September 30, 2026
Last Sunday, Apollo’s chief economist, Torsten Slok, circulated a research note suggesting that AI agents like Meta’s Muse have the potential to topple banks by making it easy for cash to flood out of those paying low yield and into other banks willing to pay higher yield. It makes for a good headline, but it doesn’t align with the reality of how humans actually behave.
It’s an appealing thought, though. The $19.6 trillion market for bank deposits in the U.S. is grossly inefficient. There’s a huge supply-demand mismatch among banks, resulting in more than 400 basis points (4.00%) of rate disparity between the lowest- and highest-yielding savings accounts. While most financial advisors think about equities and fixed income and alternatives when allocating assets, one of the greatest opportunities for generating excess risk-free returns exists in a sleepy corner of the market that gets little attention: cash. Many don’t even think of cash as an asset class, even though — according to Capgemini’s 2026 World Wealth Report— it comprises 24% of high net worth households’ assets.
This was my thesis 13 years ago when I left Citigroup to found Max, an automated cash management solution designed to help affluent households maximize both FDIC insurance coverage and yield without the risks inherent in legacy cash sweep programs. It was a novel concept: could technology help equalize this massive and inefficient market by helping people move money among their own FDIC-insured bank accounts in search of the best yield each month?
As it turns out, achieving this ideal is harder than it looks. Solving this market inefficiency is in part a technology problem, true, but it’s largely a behavioral-economics problem. When offered a free lunch, will depositors choose to nourish themselves, or continue to starve? Up until now, most depositors remain starved for yield.
When we set out address this problem in 2013, we quickly figured out how to do what AI promises to do today. That was relatively easy. Then we had to figure out how to access accounts even when the user wasn’t present. That was harder, and so we patented it. The much larger challenge, however, was how to change client behavior, and earn their trust in the process. It’s the trust piece that’s most important. After all, cash is the asset class that you hold for safety and liquidity. Rightly, no one wants to take risk with their cash.
The approach to cash management being pitched by AI companies is that you should share your bank passwords with them and trust that they’ll do no harm. That’s a non-starter for most people, and for good reason. If things go awry, who is to blame? Will the bank take your side? What happens when your money is wired to the wrong place? Which regulations exist to protect errant transfers? What about the risk of fraud? At Max, we spent the past decade solving these problems by collaborating with banks to build a trusted network that didn’t involve any sharing of account credentials. We put in place a system of guardrails that limited how money could move. And we left the money movement piece to the banks themselves. In 13 years of operation, we’ve never touched a single dollar of client funds. By working in partnership with banks — not against them — we were able to build a secure solution that’s trusted by banks, their clients, and financial advisors at more than 3,000 wealth management firms nationwide. We even helped prevent banks from failing during the 2023 regional banking crisis by helping them attract deposits when they needed them most.
And yet, we’re only scratching the surface. The vast majority of Americans still hold their cash at the largest money-center banks, where it’s often both under-insured and under-earning, often to the tune of tens of thousands of dollars per year in foregone interest income. The dearth of technological solutions for managing cash isn’t what keeps banks afloat, it’s the fact that depositors have better things to do with their time that monitor interest rates. Much to the dismay of economists, that’s a very rational sentiment. It’s also what drives profitability for the largest banks and brokerage firms, hoping that clients won’t notice their paltry yield, or more likely, won’t prioritize doing something about it.
The risk to banks then is not AI in and of itself. Far from it. The risk is that Americans will wake up and finally demand that their cash start earning more.
Gary Zimmerman is a former investment banker and the founder and CEO of MaxMyInterest, an intelligent cash management solution for individual investors and their advisors.