NEW YORK — Jamie Dimon said JPMorgan Chase is already seeing meaningful financial benefits from artificial intelligence deployments across the bank, while also warning that emerging AI models are dramatically increasing cybersecurity risks for the financial system.
Speaking about the $4.9-trillion bank’s use of AI, Dimon said JPMorgan currently has about 1,000 AI use cases underway, with roughly 50 to 60 considered “significant.”
“We’re saving real money, and we see real changes taking place,” Dimon said.
Dimon said AI is unlikely to alter the core functions of banking, such as moving money, raising capital and managing funds, but he said the technology could transform nearly every operational aspect surrounding those services.
“Everything else can change: how that gets done, how the blockchain gets used, and all these other things,” he said.
Growing Competition
Dimon also emphasized growing competition facing the banking industry from banks, nonbanks and fintech firms.
“Competitors are very smart, and they’re coming,” he said. “Some have been quite successful and have taken pieces of our business that we coulda, woulda, shoulda.”
He said one of the industry’s strongest competitive advantages is maintaining a culture focused on innovation and continued investment.
“One of the biggest moats is having a bank that is hungry and not complacent and not arrogant and constantly investing in its future, like technology,” Dimon said.
The View on Payments
In payments, Dimon said JPMorgan’s scale and global reach alone will not guarantee future dominance, particularly as stablecoins and tokenized deposit products gain traction.
“If we don’t build a new set of things, even for stablecoins or JPMorgan deposit coin, that could be challenged, too,” he said. “We are hyper-focused on all these forms of competition and constantly investing to compete in that world.”
Dimon said the bank is beta-testing Smart Cash, an AI-powered tool designed to automatically shift customer funds between checking accounts and higher-yield brokerage products to maximize returns and improve cash-flow management.
He also warned that advances in AI are heightening cybersecurity threats across the banking system, specifically citing Anthropic’s Mythos AI model.
“Mythos just amplifies it, dramatically,” Dimon said.
A ‘Nuclear Weapon’
In April, Scott Bessent and former Jerome Powell reportedly warned major bank CEOs about cybersecurity concerns tied to the model.
Dimon said Anthropic acted responsibly by alerting government officials and launching Project Glasswing, an initiative focused on safeguards and risk controls surrounding the model. JPMorgan is participating in the effort, he said.
“All the big banks are working together now,” Dimon said. “We’re trying to inform other banks where we are.”
Dimon said broader access to Mythos may eventually be appropriate but cautioned that any rollout must be carefully controlled.
“It is dangerous,” he said. “It’s a nuclear weapon in the hands of someone.”
He added that government officials and Anthropic will ultimately determine how and when wider access to the technology is granted.
