Wall Street is cutting ordinary roles while paying up for AI talent, and the hiring pattern shows banks want to own the systems reshaping their cost base.
The AI talent war on Wall Street is no longer an abstract argument about future productivity. In a Bloomberg interview aired in May, Jamie Dimon said JPMorgan would hire more AI people and probably fewer bankers in some categories. Business Insider, reporting on the interview, noted that JPMorgan has more than 300,000 employees and sees roughly 10% annual attrition, which gives the bank a large lever for changing the shape of its workforce without announcing a blunt round of AI layoffs.
The money behind that shift is already visible. Business Insider separately reported that JPMorgan set a $19.8 billion technology budget for 2026, up nearly 10% from the prior year, with about $1.2 billion directed toward investments that include AI-related projects. CFO Jeremy Barnum told investors the bank was focusing on high-impact uses such as call centers, client insights and tools for software engineers. Those are not side experiments. They sit close to the daily machinery of the bank.
That is why the talent side matters. Banks can buy plenty of AI software from vendors, and many will. But the largest firms are also trying to hire the people who know how to turn models, data controls and internal workflows into systems a regulated bank can actually use. Business Insider reported in November that six of the largest US banks, including JPMorgan, Goldman Sachs and Morgan Stanley, had posted more than 2,000 AI roles over the previous 12 months, based on Revelio Labs data. The same report said average finance salaries for AI professionals, excluding C-suite roles, had climbed to about $180,000 from $142,000 in 2020.
The profile of the hires tells you more than the job-posting count. Citi named Shobhit Varshney, a longtime IBM executive, as its head of AI in September 2025, according to Business Insider, with Varshney reporting to Citi chief operating officer Anand Selva and working with chief technology officer David Griffiths. At IBM, Varshney led the data, AI and automation business for IBM Consulting across the US, Canada and Latin America. That is a different hire from a bank bringing in a vendor manager to run procurement.
Citi also gives a useful view of how broad this work is becoming inside banks. Business Insider reported in January that Citi had built an internal AI Champions and Accelerators program with around 4,000 voluntary participants, while its AI tools were available in 84 countries for 182,000 employees. Some of that is training and adoption work, not pure engineering. Still, it shows where the large banks are heading: AI is being pushed into operations, compliance, customer service and technology teams at the same time.
If you are an AI startup selling into finance, this changes the pitch. Banks remain attractive customers because they have large budgets, messy data and thousands of repetitive workflows. But the biggest institutions are not waiting passively for outside tools to define the future of their businesses. They are hiring senior operators, training internal staff and deciding which parts of the stack they want to control themselves. The best opportunity may be in the gaps a bank cannot build quickly, not in promising to replace the systems the bank is already trying to own.
The jobs math is harder to ignore
There is also a harder workforce story beneath the hiring. The Wall Street Journal reported in January that Wells Fargo CEO Charlie Scharf expected AI's impact on staffing levels to be extremely significant, although it could take years to play out. Wells Fargo's workforce had fallen from about 275,000 people in 2019 to roughly 210,000, and Scharf said the bank would retrain workers while using attrition as its friend.
Dimon has made a similar point in less dramatic language. He told Bloomberg that AI would affect every job, and that JPMorgan would reskill displaced workers, offer other roles or potentially use early retirement. That sounds humane compared with a mass layoff announcement, but it still changes the bargain for employees. A role that is not replaced after attrition disappears just as surely as one cut in a restructuring memo.
The tension is the story. Banks are not simply using AI to trim expenses, and they are not simply hiring technologists because every board wants to hear the word AI. They are trying to concentrate control over automation in-house, close to the people who understand their data, risk systems and regulators. For startups, workers and smaller financial firms, that is the part worth watching. The largest banks are not only buying the next wave of financial technology. They are trying to build enough of it that they do not have to wait for anyone else.
Also read: Chinese AI Models Are Learning to Game Safety Tests and No One is Ready • Anthropic CEO called for AI oversight powers and the US just used them on him • India is turning its AI pitch into an export strategy
