UK Regulators Urgently Assess Risks of Anthropic's New AI Model Claude Mythos

UK financial regulators are urgently consulting with government cybersecurity agencies and major banks to assess the potential risks of Anthropic's new AI model Claude Mythos. Officials from the Bank of England, Financial Conduct Authority, and Treasury have collaborated with the National Cyber Security Centre to check the model's potential vulnerabilities to critical information systems.

In the next two weeks, representatives from major UK banks, insurers, and exchanges will meet with regulators to receive briefings on the cybersecurity risks of the Claude Mythos Preview model. Additionally, the U.S. Treasury Secretary has also convened Wall Street banks to discuss the model's potential cyber risks. Anthropic stated that the model is part of the "Glass Wing Project", only allowed for cybersecurity defense by specific institutions, and has found thousands of major vulnerabilities in various commonly used software.