Fed Chair Jerome Powell, Treasury Secretary Scott Bessent Meet with Bank CEOs to Address Cybersecurity Risks of Anthropic’s AI Model Mythos
April 10, 2026 – Washington, D.C. — Federal Reserve Chair Jerome Powell and Treasury Secretary Scott Bessent convened a high-level, closed-door meeting on Tuesday with top bank executives to discuss cybersecurity concerns linked to Anthropic’s latest artificial intelligence (AI) model, Mythos. Sources familiar with the discussions informed CBS News that the meeting took place at the Treasury Department’s headquarters in Washington.
The gathering was called amid increasing scrutiny on AI’s potential to introduce new security vulnerabilities in key financial systems and infrastructure. Bloomberg News earlier reported on the meeting, noting that several leading financial institutions were represented, though JPMorgan Chase CEO Jamie Dimon was invited but did not attend.
Anthropic, a prominent developer of generative AI technologies and the creators of the chatbot Claude, recently announced an initiative to collaborate with major technology players including Amazon, Apple, and Nvidia. This collective project will leverage Mythos to bolster cybersecurity defenses across sectors. However, Anthropic also revealed that it will not release Mythos widely to the public given its advanced capabilities— which have already exposed vulnerabilities within major operating systems and popular web browsers.
“Given the rapid pace of AI advancement, it will soon be possible for similar powerful capabilities to spread beyond those committed to deploying them responsibly,” Anthropic stated in an official post about the initiative. “The potential fallout—for economies, public safety, and national security—could be severe.”
The effort has been named Project Glasswing and is described as an urgent attempt to apply Mythos’s capabilities strictly for defensive purposes.
A Treasury Department spokesperson emphasized the government’s proactive stance on emerging AI threats, stating, “The White House has been leading an ongoing core interagency taskforce, which includes the Treasury, that has been actively engaging industry and other government bodies to implement early phases of a plan to safeguard the United States and its citizens.” The spokesperson also indicated plans for continued coordination meetings involving regulators and financial institutions in the weeks and months ahead to stay ahead of evolving AI risks.
The Federal Reserve declined to comment on the meeting.
This development follows the Biden Administration’s 2023 recognition of artificial intelligence as a potential risk to financial stability—marking the first time AI was formally designated as such by federal authorities, as previously reported by CBS News.
The dialogue between regulators and financial leaders underscores the growing concern within government circles about the dual-use nature of advanced AI technologies—capable of both strengthening defenses and potentially exposing systemic vulnerabilities if misused.
About Anthropic and Mythos
Anthropic has gained prominence for its work in building sophisticated AI language models and chatbots aimed at enhancing human-computer interaction. Mythos represents the company’s latest breakthrough model, distinguished by its capability to detect and analyze cybersecurity weaknesses with unprecedented precision. While Anthropic is advancing efforts to channel Mythos towards defensive applications, its decision to restrict widespread release reflects broader fears about the implications of unregulated access to such powerful AI tools.
Looking Ahead
As AI continues to reshape the technological landscape, the Treasury and Federal Reserve’s engagement with leading financial institutions signals an urgency to establish robust safeguards. The financial sector’s stability depends increasingly on managing digital risks posed by novel AI capabilities. With Project Glasswing and ongoing interagency coordination, the government aims to create a resilient framework addressing both current and future cybersecurity challenges fomented by sophisticated AI technologies.
Contributing Reporters: Aimee Picchi, Associate Managing Editor, CBS MoneyWatch; Richard Escobedo, Economic Policy Correspondent, CBS News
For further updates on AI and financial cybersecurity, stay tuned to CBS News.
© 2026 CBS Interactive Inc. All rights reserved.