US Treasury and Fed summon bank CEOs over Anthropic AI cyber risk
TL;DR: Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell called Wall Street’s top executives to an urgent meeting at Treasury headquarters, warning them about cybersecurity risks posed by Anthropic’s new Claude Mythos model. The gathering signals that AI safety concerns have escalated from technical discussions to financial system-level responses — a shift that UK financial regulators and City institutions will be watching closely.
Bloomberg reports that Bessent and Powell assembled the group on Tuesday to ensure banks are aware of the risks raised by Mythos and similar future models, and are taking precautions to defend their systems. The meeting was held privately, with attendees speaking on condition of anonymity.
From AI labs to financial regulators
The urgency of the meeting reflects the speed at which Anthropic’s Mythos capabilities have moved from an internal research finding to a matter of national financial security. Earlier this week, Anthropic announced Project Glasswing — a defensive cybersecurity initiative sharing Mythos with 12 major tech firms and over 40 additional organisations after discovering the model could outperform most humans at exploiting software vulnerabilities.
That both the Treasury Secretary and the Fed Chair personally convened bank CEOs suggests the US government views AI-enhanced cyber threats as a systemic risk to financial stability, not merely an IT concern. This represents a notable escalation: previous government engagement with AI risks has typically involved agency-level working groups rather than direct calls from the most senior financial officials.
Looking forward
For UK financial institutions, the implications are direct. The Bank of England and FCA have been expanding their own focus on AI-related operational resilience, but have not yet convened an equivalent high-level response. As frontier AI models grow more capable at identifying financial system vulnerabilities, the gap between US and UK regulatory urgency may become harder to sustain. The Bessent-Powell meeting sets a benchmark for how seriously governments treat AI as a financial stability risk.