JPMorgan CEO Jamie Dimon said on Tuesday that while artificial intelligence tools may ultimately help companies defend themselves against cyberattacks, they initially make them more vulnerable. Dimon said JPMorgan is testing a preview version of Mythos, Anthropic's latest artificial intelligence model released last week, as part of a broader effort by the firm to harness the benefits of artificial intelligence and prevent malicious actors from exploiting the technology.
"Artificial intelligence makes it worse and more difficult," Dimon told analysts on a JPMorgan Chase earnings call on Tuesday morning. "It does bring additional vulnerabilities, but maybe in the long run, it also brings better ways to strengthen your own defenses."
When asked about the Mythos model, Dimon said: "I think you've read about it, it shows that there are more vulnerabilities that need to be fixed." This seemed to refer to Anthropic's previous warning that the model has found thousands of vulnerabilities in enterprise software.
The comments reveal how artificial intelligence, a technology that businesses viewed as a productivity boon, has turned into a serious threat as it provides bad actors with new ways to hack into technology systems. Last week, U.S. Treasury Secretary Scott Bessant convened a meeting of major bank CEOs to discuss the risks posed by the Mythos vulnerability.
Dimon said that as the world's largest bank by market capitalization, JPMorgan Chase has invested heavily over the years to deal with various threats, set up dedicated teams and maintained close coordination with government agencies.
"We have invested a lot of money. We have top experts. We are in close contact with the government," he said. “It’s a full-time job and we’re constantly working on it.”
However, given the interconnectedness of the financial system, the CEO warned that the risks do not reside solely within any single institution.
"That doesn't mean everything that banks rely on is tightly protected," Dimon said. "Banks...are connected to exchanges and various other institutions, and those create other layers of risk."
Dimon also said that while advanced artificial intelligence tools are important, traditional cybersecurity measures remain critical.
"A lot of times, cybersecurity falls into the realm of hygiene... How do you protect your data? How do you protect your network, your routers, your hardware? How do you change your passwords regularly?" he said. "Doing these things can significantly reduce risk."