Powell, Bessent Warn Banks About Security Risks From Anthropic’s Mythos AI: Bloomberg

US Treasury Secretary Scott Bessent and Federal Reserve Chairman Jerome Powell reportedly convened a meeting with Wall Street bank CEOs earlier this week to warn about the cybersecurity risks associated with a new artificial intelligence model from Anthropic.

According to one report Through BloombergThe meeting included executives from Citigroup, Bank of America, Wells Fargo, Morgan Stanley and Goldman Sachs. Officials discussed Anthropic’s new AI model Mythos, which has recently raised widespread concerns about its apparently advanced cybersecurity capabilities.

Officials convened the meeting to ensure banks understand the risks associated with systems that can identify and exploit software vulnerabilities in operating systems and web browsers, and to encourage institutions to strengthen their defenses against potential AI-enabled cyberattacks on financial infrastructure.

Security researchers have warned that tools capable of automatically discovering vulnerabilities could accelerate both defensive security work and malicious hacking if exploited.

Anthropic’s Mythos model first appeared online in March after draft material about the system was leaked online, revealing what the company described as its most capable AI model yet. During testing, the system reportedly found thousands of previously unknown software vulnerabilities, including zero-day flaws in major operating systems and web browsers.

Anthropic researchers said in a report earlier this week, it became clear that Mythos Preview’s vulnerability discovery capabilities were not intentionally trained, but instead emerged from broader improvements in the model’s coding, reasoning, and autonomy.

“The same improvements that make the model significantly more effective at patching vulnerabilities also make it significantly more effective at exploiting them,” the company wrote.

Because of these capabilities, Anthropic has limited access to a small group of cybersecurity organizations.

“Given the strength of its capabilities, we are being deliberate about how we release it,” Anthropic said in a statement. “As standard industry practice, we are working with a small group of early access customers to test the model. We view this model as a step change and the most capable we have built to date.”

To address that risk, Anthropic is testing Mythos through Project Glasswing, a partnership with major technology and cybersecurity companies that use the model to identify and patch vulnerabilities in critical software before attackers can exploit them.

“Project Glasswing is a starting point. No one organization can solve these cybersecurity challenges alone,” the company said in a press release statement. “Frontier AI developers, other software companies, security researchers, open source maintainers, and governments around the world all have a vital role to play.”

Anthropic did not immediately respond Declutter request for comment.