Banks could face significant challenges as advances in artificial intelligence (AI) enable customers to better deploy funds traditionally left idle in deposit accounts. This development may disrupt the longstanding model in which banks profit from holding low- or non-interest-bearing deposits and investing them at higher rates.

Agentic AI assistants, such as Meta’s Muse, are designed to help savers identify more advantageous uses for surplus balances. These AI tools could range from simply alerting customers to better interest offerings to automatically transferring excess funds into higher-yielding money market accounts. Such shifts have the potential to reduce the volume of deposits held at minimal or zero interest, which currently constitute a substantial portion of banks’ funding bases.

Major US banks like JPMorgan Chase, Bank of America, and Wells Fargo collectively hold around $1.6 trillion in non-interest-bearing deposits, representing approximately 16% of their liabilities. Analysts estimate that if these funds were instead paid a market rate of around 3%, the banks could face an annual cost increase of $47 billion—nearly half of their combined annual profits. Expanding this calculation across the entire US banking sector, where deposits total roughly $20 trillion with about 80% classified as “core” deposits, the potential equity value at risk is estimated at $500 billion. This figure equates to roughly 10% of the market capitalization of listed US banks.

However, experts caution against exaggerated predictions of profit erosion or “agentic bank runs,” noting several factors that could limit the pace of fund movement. A key obstacle is trust; customers may be hesitant to relinquish control of their funds to AI entities perceived as risk-prone or error-prone. Moreover, operational frictions and regulatory considerations may slow the advance of fully automated fund reallocations.

While banks may lose revenue on deposits, they stand to gain from AI-enabled efficiencies in other areas. These include reduced operational costs, improved loan pricing, and mitigation of costly human mistakes in transactions. Bank of America’s CEO Brian Moynihan recently reported that an investment of $400 million in AI initiatives yielded an estimated $800 million in benefits for the institution. Such gains could partly offset losses from shrinking spreads and may also enhance returns to depositors.

Despite these emerging AI-driven trends, many investors remain more concerned about traditional risks, particularly the impact of ongoing Federal Reserve interest rate hikes on bank valuations. A recent survey by lender Truist found that more than half of bank investors identified rate increases as the primary threat, with relatively few viewing AI-induced disruptions as significant risks. JPMorgan CEO Jamie Dimon has similarly highlighted that the banking sector faces larger challenges related to credit cycles, with a major downturn overdue after more than a decade of relative stability.

In summary, while AI technologies are poised to reshape how consumers manage idle funds and could pressure banks’ deposit-based revenues, the industry’s overall value and stability hinge on a broader set of economic and operational factors.