Banks Aren’t Laying People Off. They’re Just Not Rehiring Them

AI is changing how large U.S. banks manage staffing, but not primarily through layoffs. At Bank of America, the change is visible in how the bank decides whether to refill roles after employees leave.

Bank of America ended 2025 with roughly 213,000 employees, a figure that had remained largely unchanged for two years, and executives have said that number is expected to decline in 2026. During the bank’s fourth-quarter earnings call, chief executive Brian Moynihan said the focus is on managing expenses through operational efficiency and the use of new technology, including artificial intelligence.

The Replacement Decision has Changed

The bank has not announced large-scale layoffs. Instead, executives have described a process in which departures trigger a review of whether a role still needs to exist. During the same earnings call, chief financial officer Alastair Borthwick said, “Every time someone leaves, we take the opportunity to evaluate whether the role needs to be replaced.”

That approach allows headcount to fall without formal job cuts. In 2025, Bank of America hired about 17,000 employees, primarily to replace workers who left, and added about 2,000 college graduates, even as executives said total staffing levels are expected to decline as roles are reassessed rather than automatically refilled.

The shift is supported by the bank’s technology deployment. Bank of America spends about $13 billion annually on technology, with roughly $4 billion allocated to new initiatives, including AI projects. More than 90% of employees use its internal AI assistant, Erica for Employees, for tasks such as IT support and internal troubleshooting. According to Bank of America, Erica for Employees has reduced internal IT service-desk calls by about 50%.

During the question-and-answer portion of the earnings call, analysts asked how increased technology spending would affect expenses and staffing. Management said overall technology spending remains around $13 billion, with about $4 billion directed toward new initiatives, and described “several hundred million dollars” allocated to AI across roughly 15 to 20 active projects. Moynihan also said AI techniques have reduced about 30% of the coding workload, which he described as equivalent to the output of roughly 2,000 developers.

The bank’s financial disclosures show compensation and benefits expense of about $40.18 billion for 2025, up from roughly $38.33 billion in 2024, alongside technology spending of about $13 billion. Executives have said they are directing a portion of the technology budget toward AI while monitoring compensation and other non-interest expense categories.

Where Staffing Pressure is Concentrated

Executives have pointed to specific functions where AI affects staffing needs. Moynihan cited the bank’s audit team, which expanded in recent years to manage increased regulatory requirements, as an example where AI-powered tools now allow audit work to be completed with fewer staff while maintaining oversight.

Productivity gains in software development have also been cited. Moynihan said AI coding tools now provide output equivalent to the work of about 2,000 developers. Bank of America has not announced layoffs of that size in its technology organization. Instead, the productivity gains allow existing teams to take on additional work without expanding headcount.

At the same time, the bank continues to hire in areas tied to customer-facing roles and early-career recruitment. The workforce is changing unevenly, with reductions concentrated in internal and process-driven functions rather than across all job categories.

Similar patterns are visible across the banking sector. The six largest U.S. banks reported a combined workforce of about 1.09 million employees at the end of 2025, down roughly 10,600 from a year earlier, according to data compiled by Bloomberg. That total is the lowest since 2021.

Other banks are using different approaches. Citigroup is carrying out a multi-year restructuring that includes significant workforce reductions linked to operational simplification. Wells Fargo reported about $612 million in severance expenses in the fourth quarter of 2025, and its chief executive, Charlie Scharf, has said artificial intelligence will affect staffing levels over time.

Bank of America’s approach relies more heavily on attrition and role evaluation. The bank continues to hire selectively but does not assume that every departure requires a replacement. AI tools reduce the volume of work associated with some roles, which changes how staffing decisions are made.

The result is a gradual decline in headcount without a single workforce action or announcement. The change is occurring through replacement decisions rather than layoffs.