AI Could Hit Borrowers Before It Hits Banks, Says JPMorgan
Highly leveraged software firms face rising borrowing costs as AI reshapes their markets
Bank stocks fell this week after a research note suggested artificial intelligence could change how everyday payments work. Shares of JPMorgan Chase dropped more than 4% alongside other financial firms.
At an investor event in Manhattan, Chief Executive Jamie Dimon pushed back. “In my view, we’ll be a winner,” he said. The bank plans to spend about $19.8 billion on technology this year.
But during the same event, Dimon warned about lending. He said some competitors are doing “dumb things” to boost net interest income. He also said that in the next credit downturn, software could be a surprise area of weakness.
AI may help banks improve how they operate. It may also affect the companies that owe them money.
AI is changing how banks run
JPMorgan’s technology budget is one of the largest in banking. The bank expects to spend about $19.8 billion on technology in 2026. Chief Financial Officer Jeremy Barnum said machine learning and analytical AI have been driving improvements in revenue and expense for years
AI tools are now used across lending decisions, pricing and fraud detection. In the fourth quarter of 2025, JPMorgan’s payments unit generated about $5.1 billion in revenue out of roughly $46.8 billion in total managed revenue for the quarter, according to the bank’s earnings release.
Dimon acknowledged that competition in payments has intensified. He named fintech rivals including Revolut, PayPaland Stripe. “We got beat badly,” he said according to The Wall Street Journal.
The bank has used AI internally for years. One system, known as COiN, was designed to review commercial loan contracts and reduced work equivalent to about 360,000 hours a year. More recently, its asset and wealth management arm introduced an internal AI system to analyze proxy votes rather than relying on outside advisory firms.
AI may also pressure borrowers
Dimon’s warning about software points to the other side of the balance sheet.
Software and technology borrowers account for roughly 17% of the U.S. leveraged loan market, with software loans totaling about $260 billion. About half of those borrowers are rated B- or lower, meaning they already carry higher credit risk.
In recent months, lenders have demanded higher yields and stricter terms for some software borrowers. Several companies have delayed debt deals as investors reassess how AI could affect future earnings.
Ratings agency Morningstar DBRS reported that downgrades in private credit now outnumber upgrades by more than three to one. The share of riskier credits has risen to 16%, and default rates have climbed to around 4%.
Private credit has grown into a multi-trillion-dollar market over the past decade.
If AI increases competition or reduces pricing power for certain software companies, profits could fall. When profits fall, debt becomes harder to repay.
JPMorgan has said it expects card net charge-offs of about 3.4% and has described consumer spending as steady across income groups.
AI may help banks operate faster and more efficiently. At the same time, it could reshape the industries that borrowed heavily in recent years. That makes AI not just a technology story for banks, but a credit story as well.