Jamie Dimon is worried about escalating cyber risks due artificial intelligence, all while the bank he oversees, JPMorgan Chase & Co., continues to finance.
The bank boss pointed to Anthropic and its Mythos AI model, emphasizing growing concerns about advanced AI systems taking unintended actions online.
"Risks from AI went up 10-fold after Mythos," Dimon told Bloomberg on Tuesday. AI has introduced new security weaknesses while businesses continue to deal with conventional cyber threats.
JPMorgan’s Ties to Anthropic
Dimon’s comments come as JPMorgan has deepened its financial ties to Anthropic and the broader AI boom.
For example, JPMorgan was among the investors in Anthropic’s $30 billion Series G funding round announced in February and has been involved in many of the AI company’s banking and credit arrangements.
JPMorgan, along with Morgan Stanley, Goldman Sachs and Citigroup, has prominent roles in Anthropic’s $15 billion pre-IPO credit facility, according to Bloomberg. The bank is also reportedly working on Anthropic’s planned initial public offering.
At the same time, JPMorgan is one of the launch partners in Anthropic’s Project Glasswing, giving it access to the company’s Mythos AI system to test software for vulnerabilities.
Incidents Are Piling Up
Earlier this year, Anthropic found three incidents in which a Claude model accessed the internet unauthorized while interacting with a third-party evaluation environment, the company revealed in a press release at the time.
"After reviewing 141,006 evaluation runs where Claude could have obtained internet access, we identified three incidents in which a model accessed the internet from within or while interacting with the evaluation environment of Irregular, one of our third-party evaluation partners, and then gained unauthorized access to the production infrastructure of three different organizations."
The cybersecurity incident raised concerns regarding AI systems carrying out tasks in unexpected ways, including attempts to modify online software with potentially harmful code.
OpenAI disclosed that one of its most advanced autonomous AI agents broke out of a highly isolated testing environment, accessed the internet, and hacked AI platform Hugging Face while attempting to complete its assigned evaluation objective.
Dimon said the risks associated with AI agents and systems such as Mythos are legitimate but argued that the focus should remain on addressing them rather than get hung up on whether they pose an existential threat.
Dimon Is Not Getting ‘Hysterical’
"I’m not going to get hysterical over, ‘Is it existential or not?’ What we’re doing is rolling up our sleeves and going to work to fix it," Dimon told Bloomberg.
Dimon also argued that data centers should be developed in communities that are willing to support them.
"There are a lot of states that would take it that have access to power," the CEO said.
Dimon previously made comments surrounding AI and Anthropic’s Mythos AI model, warning that the risks stemming from Mythos were "a real issue." He noted that U.S. government officials are now focused on addressing the potential risks.
Dimon also warned that giving widespread access to Mythos could create risks comparable to “giving ballistic missiles to individuals,” the CEO said during Republican Sen. Dave McCormick’s Pennsylvania Defense and Innovation Summit.
Photo: Shutterstock