The Federal Reserve convened an extraordinary meeting in April 2026 with the CEOs of the nation's top banks and Treasury Secretary Scott Bessent to warn about the cybersecurity risks posed by Anthropic's advanced artificial intelligence model, Claude Mythos Preview. The Fed highlighted that the model could present unprecedented threats to the nation's top financial institutions, given its ability to identify weaknesses and security vulnerabilities within software [1]. Anthropic released Claude Mythos Preview to a select group of banks and institutions as part of Project Glasswing, a cybersecurity initiative [1].
Despite raising the alarm, the Federal Reserve itself lacked access to the Mythos AI model for at least three months following the meeting, leaving it potentially vulnerable while other institutions began patching their weaknesses using the tool [1]. As of July 15, 2026, the Fed was still attempting to secure access to Mythos, and it remains unclear whether it has since obtained it [1]. Anthropic did not respond to requests for comment, and the Fed also declined to comment for the article [1].
In recent congressional testimony, Fed Chairman Kevin Warsh stated he was actively working to secure access to Mythos and other cutting-edge AI models, emphasizing the need for the Federal Reserve and other institutions to protect themselves from emerging vulnerabilities. Warsh noted, "We are not the deciders as to who has access, but I have not been shy in sharing my views with authorities across the government about the vulnerabilities, and have been asking for access not just for the Federal Reserve but for other institutions to a whole range of these new artificial intelligence models so that they can protect themselves" [1]. He further clarified that the Fed's cybersecurity efforts should not be limited to Mythos alone, as broader adoption of new AI models is necessary for the banking system's security [1].
Warsh has embraced AI as a transformational technology during his tenure, but it is unclear whether the Fed can begin addressing its vulnerabilities without access to Mythos [1].
CONCLUSION
The Federal Reserve's lack of access to Anthropic's Mythos AI model, despite warning banks of its cybersecurity risks, has raised concerns about the central bank's vulnerability. While other institutions have begun patching their systems, the Fed continues to seek access, highlighting the urgency and high market impact of advanced AI in financial cybersecurity. The situation underscores the need for equitable access to cutting-edge AI tools across the financial sector.