In the wake of Anthropic's unveiling of its powerful Claude Mythos AI—capable of detecting and exploiting software vulnerabilities—the US Treasury Secretary has convened top bank executives to highlight escalating AI-driven cyber threats. The move underscores growing concerns as the AI is restricted to a tech coalition via Project Glasswing.
Following Anthropic's April 7 announcement of Claude Mythos, a general AI model that surpasses most humans in coding and vulnerability exploitation (including a 27-year-old bug in OpenBSD), Treasury Secretary Scott Bessent this week summoned executives from Bank of America, Citigroup, Goldman Sachs, Morgan Stanley, and Wells Fargo, alongside Fed Chair Jay Powell. Sources cited by the Financial Times report Bessent warned of heightened risks from AI-enhanced cyberattacks.
JPMorgan Chase CEO Jamie Dimon, who was absent, had previously highlighted how AI exacerbates cyber threats. The meeting reflects broader industry alarm over Anthropic's decision to withhold public release of Mythos, instead channeling it through Project Glasswing—a coalition of around 40 tech organizations including Apple, Amazon, Microsoft, Google, and the Linux Foundation—to patch flaws responsibly.
Skepticism persists, however. Analyst Gary Marcus labeled the announcement 'inflated marketing hype,' pointing out that Firefox tests disabled sandboxing and that smaller open-source models already perform similar vulnerability analysis. This follows Anthropic's initial reveal, which detailed thousands of flaws found in major OSes and browsers.
