British financial regulators are in urgent discussions with banks and cybersecurity officials to examine the potential risks associated with Anthropic's latest artificial intelligence model, 'Claude Mythos.' Launched recently for cybersecurity applications, the AI model has triggered significant concern among government officials in both the United Kingdom and the United States.
Urgent Regulatory Scrutiny
Key UK financial authorities, including the Bank of England, the Financial Conduct Authority (FCA), and the Treasury, are collaborating with the National Cyber Security Centre (NCSC). Their objective is to determine whether Anthropic's new AI model could pose threats to critical IT systems within the finance sector.
Regulators are preparing to brief banks, insurers, and exchanges within the next two weeks on the cybersecurity risks linked to Claude Mythos. Anthropic introduced the model as part of its 'Project Glasswing' initiative, aiming to deploy it across enterprise and critical infrastructure use cases through partnerships with major companies like Google and Apple, as well as financial firms.
Dual-Use Concerns and US Response
While Anthropic claims Mythos has already identified thousands of zero-day vulnerabilities and decades-old flaws, the model's capabilities have also raised alarms. Officials are particularly concerned about the dual-use nature of the technology, fearing it could be exploited by malicious actors for offensive cyber operations.
In the United States, similar concerns prompted Treasury Secretary Scott Bessent to hold an urgent meeting with major Wall Street banks. Executives from institutions such as JPMorgan, Bank of America, Citigroup, Goldman Sachs, Morgan Stanley, and Wells Fargo reportedly attended to discuss the model's cyber risk potential.
Anthropic's Broader Challenges
Beyond the regulatory reviews, Anthropic is also embroiled in a legal dispute with the US Trump Administration and the Pentagon. The company is challenging its blacklisting and designation as a 'supply chain risk,' a label that a Washington, D.C. federal court recently declined to remove.