Anthropic’s new AI model, Claude Mythos Preview, has drawn close attention from U.S. financial officials after the company limited access to a small group of firms. The model was introduced through Anthropic’s Project Glasswing program, which focuses on defensive cybersecurity work. U.S. officials then held meetings with banks and technology leaders to review potential cyber risks associated with advanced AI systems.
US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell met with major bank executives in Washington this week to discuss cyber threats tied to advanced AI tools. The meeting focused on how banks should prepare for systems capable of detecting software weaknesses at high speed.
Kevin Hassett, director of the White House National Economic Council, later said Bessent and Powell went through the cyber risks with bank leaders so they understood the issue clearly. The session showed that financial authorities are treating AI-related cyber threats as an urgent matter.
The concern centers on the banking system’s heavy dependence on software, digital records, and networked operations. Any weakness in those systems can create operational problems, service outages, or risks to customer accounts if attackers gain an advantage.
Anthropic launched Claude Mythos Preview on April 7 as a restricted release instead of a broad public rollout. The company said it made that choice because the model has advanced cyber capabilities and needs tighter controls.
According to Anthropic, Mythos Preview has already identified many severe vulnerabilities, including flaws affecting major operating systems and web browsers. The company said the model is designed to help trusted partners find security gaps before attackers exploit them.
Anthropic also said it had ongoing discussions with the U.S. government before the release. The company described the model’s “offensive and defensive cyber capabilities” as a reason for the limited rollout. “Offensive and defensive cyber capabilities” became a central part of the government’s review before and after the launch.
Also Read: Anthropic Hits $30 Billion Revenue, Edges Past OpenAI
Before the public rollout, Vice President JD Vance and Bessent questioned major technology executives about AI model security and how companies should respond to cyberattacks. The reported participants included leaders from Anthropic, Alphabet, OpenAI, Microsoft, Palo Alto Networks, and CrowdStrike.
Anthropic said Project Glasswing includes launch partners such as Amazon Web Services, Apple, Cisco, CrowdStrike, Google, JPMorganChase, Microsoft, Nvidia, and Palo Alto Networks. The company also said more than 40 additional groups that build or support critical software infrastructure received access.
The restricted release shows that both public officials and private firms are moving cautiously as more powerful AI systems enter the cybersecurity field. For banks, the focus remains on protecting core platforms, testing defenses, and preparing for faster, more capable digital threats.