Anthropic to Brief Financial Regulators on AI Found Banking Cyber Flaws

The artificial intelligence company Anthropic has agreed to brief global financial regulators on cybersecurity vulnerabilities in the international banking system identified by its latest AI model, following a request from the chair of the Financial Stability Board (FSB). The briefings will be presented to members of the FSB, an international body that coordinates financial rules for G20 economies and includes finance ministries, central banks, and securities regulators from major nations.

The request came from Bank of England Governor Andrew Bailey, who chairs the FSB. Bailey has publicly expressed concern about the capabilities of Anthropic's model, named Mythos, stating that it could potentially "crack the whole cyber-risk world open" by identifying system vulnerabilities that could be exploited for attacks. The planned discussions signal that financial stability authorities are treating advanced AI security research as a critical input for understanding systemic risks, not merely as a technical tool.

Anthropic unveiled the Mythos Preview model last month, describing it as a specialized AI system built to detect long-standing, high-severity vulnerabilities in software, web...