Mythos’ AI scare is actual — sufficient for U.S. regulators to name an urgent assembly and assess what Anthropic’s superior synthetic intelligence mannequin it may imply for banks.
The assembly occurred Tuesday, with Treasury Secretary Scott Bessent and Fed Chair Jerome Powell sitting down with Wall Street bank CEOs to talk about attainable cybersecurity dangers linked to Mythos, individuals aware of the matter informed Bloomberg.
Participants included chief executives from Citigroup Inc, Morgan Stanley, Bank of America Corp.’, Wells Fargo & Co.’s, and Goldman Sachs Group Inc.’s. All these are designated as systemically essential, which means disruptions to their operations may have international repercussions.
Mythos, a sophisticated synthetic intelligence mannequin developed by Anthropic, is designed to determine and exploit vulnerabilities in software program techniques when prompted. Unlike typical consumer-facing AI instruments, Mythos is geared towards cybersecurity software program engineering and cybersecurity duties. Its specialty is figuring out crucial software program vulnerabilities and bugs, however it could additionally assemble refined exploits.
The episode highlights a elementary change in how regulators are framing AI danger, not merely as a technological problem, however as a possible catalyst for systemic occasions.
This has already raised pink flags in crypto, the place specialists are worried that Mythos’ functionality of discovering and exploiting zero-day vulnerabilities in real-time at a low value poses danger to the DeFi infrastructure.
Anthropic, due to this fact, has taken a cautious strategy, releasing the product solely for small group of enormous expertise and monetary corporations beneath “Project Glasswing.”
Anthropic has beforehand disclosed that it consulted with U.S. officers forward of Mythos’ launch relating to each its defensive and offensive cyber capabilities. The firm can be individually engaged in a authorized dispute with the Pentagon, which has designated it a supply-chain danger — a classification Anthropic is contesting in court docket.
