Mythos AI threat prompts Bessent, Powell to convene bank CEOs for urgent talks

Mythos AI threat prompts Bessent, Powell to convene bank CEOs for urgent talks

Mythos’ AI scare is real — enough for U.S. regulators to call an urgent meeting and assess what Anthropic's advanced artificial intelligence model it could mean for banks. The meeting happened Tuesday, with Treasury Secretary Scott Bessent and Fed Chair Jerome Powell sitting down with Wall Street bank CEOs to discuss possible cybersecurity risks linked to Mythos, people familiar with the matter told 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 important, meaning disruptions to their operations could have global repercussions. Mythos, an advanced artificial intelligence model developed by Anthropic, is designed to identify and exploit vulnerabilities in software systems when prompted. Unlike typical consumer-facing AI tools, Mythos is geared toward cybersecurity software engineering and cybersecurity tasks. Its specialty is identifying critical software vulnerabilities and bugs, but it can also assemble sophisticated exploits. The episode highlights a fundamental change in how regulators are framing AI risk, not merely as a technological challenge, but as a potential catalyst for systemic events. This has already raised red flags in crypto, where experts are worried that Mythos' capability of discovering and exploiting zero-day vulnerabilities in real-time at a low cost poses risk to the DeFi infrastructure. Anthropic, therefore, has taken a cautious approach, releasing the product only for small group of large technology and financial firms under “Project Glasswing.” Anthropic has previously disclosed that it consulted with U.S. officials ahead of Mythos’ release regarding both its defensive and offensive cyber capabilities. The company is also separately engaged in a legal dispute with the Pentagon, which has designated it a supply-chain risk — a classification Anthropic is contesting in court. As blockchain adoption scales, the metadata available to machine learning models scales with it. Obfuscation-based privacy approaches are structurally degrading as a result. This report provides a comprehensive comparison of all five major crypto privacy architectures and a framework for evaluating which models remain durable as AI capabilities improve. Disclosure & Polices: CoinDesk is an award-winning media outlet that covers the cryptocurrency industry. Its journalists abide by a strict set of editorial policies. CoinDesk has adopted a set of principles aimed at ensuring the integrity, editorial independence and freedom from bias of its publications. CoinDesk is part of Bullish (NYSE:BLSH), an institutionally focused global digital asset platform that provides market infrastructure and information services. Bullish owns and invests in digital asset businesses and digital assets and CoinDesk employees, including journalists, may receive Bullish equity-based compensation.