Powell, Bessent discussed Anthropic's Mythos AI cyber threat with major U.S. banks
On April 10, 2026, Federal Reserve Chairman Jerome Powell and Treasury Secretary Scott Bessent convened a crucial meeting with the heads of major U.S. banks to address the cyber threats posed by Anthropic’s new AI model, known as Mythos. This meeting highlighted the increasing concerns about the potential misuse of advanced artificial intelligence technologies in the financial sector.
Background of the Meeting
The meeting took place in Washington, D.C., where bank CEOs were already gathered for a Financial Services Forum board meeting. The urgency of the situation prompted a special gathering on Tuesday to discuss the implications of the Mythos model, which had been recently rolled out by Anthropic.
Notably, Jamie Dimon, CEO of JPMorgan Chase, was the only major banking executive unable to attend the meeting. However, other prominent figures were present, including:
- Brian Moynihan – CEO of Bank of America
- Jane Fraser – CEO of Citigroup
- David Solomon – CEO of Goldman Sachs
- Ted Pick – CEO of Morgan Stanley
- Charlie Scharf – CEO of Wells Fargo
The meeting was reported by Bloomberg and the Financial Times, which indicated that the discussions were confidential and focused on the advanced capabilities of AI and their potential risks to the U.S. financial system.
Concerns Over Anthropic’s Mythos Model
Anthropic introduced the Claude Mythos Preview model in a limited capacity due to fears that hackers could exploit its advanced capabilities. This initiative is part of a broader cybersecurity effort, known as Project Glasswing, which includes partnerships with major companies such as Apple, Google, Microsoft, and Nvidia.
Before the public rollout, Anthropic briefed senior U.S. government officials on the model’s offensive and defensive cyber applications. The company has been in ongoing discussions with various government agencies, including the Cybersecurity and Infrastructure Security Agency (CISA) and the Center for AI Standards and Innovation, regarding the potential cyber capabilities of the Mythos model.
The Implications of AI in Cybersecurity
As AI technologies evolve, their implications for cybersecurity become increasingly complex. Dario Amodei, CEO of Anthropic, emphasized the importance of getting the deployment of AI technologies right, stating that there is a significant opportunity to create a more secure internet and world through AI-powered cyber capabilities. However, he also acknowledged the risks associated with misuse.
The Trump administration’s engagement with Anthropic regarding the Mythos model reflects the growing recognition of AI’s potential threats to national security. The administration has faced challenges related to the use of AI in military applications, particularly concerning Anthropic’s insistence on limiting the use of its technology in warfare.
Legal Challenges and Regulatory Scrutiny
Anthropic is currently embroiled in legal challenges concerning its classification as a supply chain risk by the Department of Defense (DOD). Recent court rulings have allowed the company to continue working with other government agencies, despite being barred from DOD contracts. This situation underscores the regulatory scrutiny that AI companies face as they navigate the intersection of technology, security, and ethics.
In March, a federal judge granted Anthropic a preliminary injunction in a separate case, allowing it to continue its operations while the legal challenges are resolved. However, the company remains under significant pressure from the government, as evidenced by the recent denial of its request to temporarily block the blacklisting.
Market Reactions and Future Outlook
The introduction of the Mythos model has already impacted the cybersecurity market. Following the revelation of its advanced capabilities, stocks in the cybersecurity sector experienced a slump. Concerns were raised after a draft blog post from Anthropic detailed the model’s potential risks, including its previous use by hackers to orchestrate AI-fueled attacks.
In November 2025, Anthropic disclosed that a Chinese hacking group had utilized its Claude model to automate attacks on government and corporate targets. This incident has intensified discussions around the ethical implications of AI and the responsibilities of tech companies in ensuring their products are not misused.
Conclusion
The discussions between Powell, Bessent, and the major U.S. bank CEOs mark a significant moment in the ongoing dialogue about the intersection of AI technology and cybersecurity. As the capabilities of AI continue to advance, the financial sector must remain vigilant against potential threats while also exploring the opportunities that these technologies present. The outcome of these discussions may shape future regulatory frameworks and industry standards for AI in the financial services sector.
Note: The information presented in this article is based on reports and statements available as of April 2026. Developments in AI technology and cybersecurity are rapidly evolving, and readers are encouraged to stay informed through reliable news sources.
