JPMorgan says AI cuts jobs in some units, but margin gains remain limited
JPMorgan is using artificial intelligence across nearly 1,000 use cases, and the technology is already reducing headcount by as much as 40% in some parts of the bank. The lender says those efficiency gains are not translating into a dramatically cheaper operating model, as competitive pressure keeps most benefits flowing to customers.
Highlights
- JPMorgan CEO Jamie Dimon states AI has reduced jobs by 30%-40% in select units, but margin expansion remains constrained as competitors adopt similar technology.
- JPMorgan posts net income of $21.2 billion in the second quarter, up 41% year over year, with investment banking fees rising 30% to $3.3 billion.
- CFO Jeremy Barnum projects token-related AI costs will remain trivial through 2026 but expects a significant increase in the second half of 2024 as usage expands.
AI deployment and cost outlook
As reported by Business Insider, CEO Jamie Dimon says during JPMorgan's second-quarter earnings call on Tuesday that AI has already reduced jobs by 30% to 40% in discrete areas of the bank, with most affected employees offered roles elsewhere.Dimon says investors should not expect AI to suddenly lower the bank's expense growth or sharply widen profit margins. He argues that in a competitive capitalist market, banks broadly adopt the same technologies, limiting any one firm's ability to retain the full financial upside.
He also says computerization over the past 20 years did not drive margins to extreme levels, and he presents AI as a similar tool that improves service rather than creating an outsized structural cost advantage for JPMorgan alone.
Revenue strength and spending implications
JPMorgan reports net income of $21.2 billion, up 41% from a year earlier, helped significantly by gains on an investment in Visa. In a press release, Dimon says each business posts record revenue in the quarter, while investment banking fees rise 30% year over year to $3.3 billion, their highest level since 2021.Chief Financial Officer Jeremy Barnum says token-related AI costs are still trivial and should remain so through the end of 2026, but the bank is forecasting meaningful acceleration in those expenses in the second half of the year. He says token spending remains an important issue as JPMorgan decides which models to use for specific tasks.
Dimon said in May that the bank will probably hire fewer bankers in certain areas and more AI specialists over time. With a technology budget of nearly $20 billion, JPMorgan is already applying AI in fraud protection, marketing and note-taking, underscoring how the bank is shifting resources even as overall operating savings remain constrained.
In our earlier coverage of Fed Chair Kevin Warsh’s testimony, we noted that the central bank is prioritizing a return to price stability while inflation remains above its 2% target. Warsh also pointed to accelerating business investment in data centers and AI equipment as a key pillar of U.S. economic resilience, even as policymakers debate whether AI-led productivity gains will ultimately be disinflationary.
- Forex
- Crypto