White House meets Anthropic CEO over Mythos AI model's security uses

White House Chief of Staff Susie Wiles met Friday with Anthropic CEO Dario Amodei to discuss the artificial intelligence company’s new Mythos model, according to the White House and Anthropic. The meeting came as the Trump administration evaluates potential national security applications of advanced artificial intelligence systems.

The encounter signals a potential shift in the administration’s approach to Anthropic, whose leadership had sought assurances the military would not use its technology in fully autonomous weapons or domestic surveillance. It underscores the government’s focus on both the security opportunities and risks posed by increasingly capable AI systems.

The meeting explored how the two could collaborate on cybersecurity, America’s position in the global AI race, and artificial intelligence safety. A White House official said the administration is broadly engaging with advanced AI labs about their models and software security, adding that any new technology the federal government might use would require a technical evaluation period.

The Mythos Model

Anthropic announced Mythos on April 7, restricting access to selected customers due to its ability to exceed human-level performance in identifying and exploiting computer security vulnerabilities. The company described the model as a significant step forward in AI capabilities and said it poses substantial security implications if misused.

The UK’s AI Security Institute evaluated the new model as a “step up” over previous iterations. “Mythos Preview can exploit systems with weak security posture, and it is likely that more models with these capabilities will be developed,” the institute said in a report.

From Conflict to Engagement

The White House meeting represented a dramatic reversal after months of public conflict. In February, President Donald Trump posted that the administration “will not do business with them again,” after Anthropic declined to remove safeguards on how the Pentagon could deploy its technology. Defense Secretary Pete Hegseth had also moved to declare Anthropic a supply chain risk—an unprecedented action against a U.S. technology company.

Anthropic challenged the declaration in federal court. The company sought assurances that the Pentagon would not use its models in fully autonomous weapons and would not employ them in surveillance of Americans. Hegseth countered that the company must permit any uses the Pentagon deemed lawful. U.S. District Judge Rita Lin ruled in March to block enforcement of Trump’s directive ordering federal agencies to stop using Anthropic products.

When asked Friday if Anthropic had held a White House meeting, Trump said he had “no idea.”

Assessing Real Risks

David Sacks, the White House’s AI and crypto czar and an influential critic of Anthropic’s public positioning, said the company’s claims about Mythos’s risks should be taken seriously. “Anytime Anthropic is scaring people, you have to ask, ‘Is this a tactic? Is this part of their Chicken Little routine? Or is it real?’” Sacks said on the “All-In” podcast. “With cyber, I actually would give them credit in this case and say this is more on the real side.”

According to Sacks, the technical progression is logical: “It just makes sense that as the coding models become more and more capable, they are more capable at finding bugs. That means they’re more capable at finding vulnerabilities. That means they’re more capable at stringing together multiple vulnerabilities and creating an exploit.”

Global Safeguards and Partnerships

Anthropic is simultaneously in talks with the European Union about its AI models, including advanced systems not yet released in Europe. European Commission spokesman Thomas Regnier confirmed Friday that discussions were underway.

To address these risks, Anthropic formed Project Glasswing, bringing together Amazon, Apple, Google, Microsoft, and JPMorgan Chase to secure critical software from potential threats posed by the new model.

Anthropic co-founder and policy chief Jack Clark said at the Semafor World Economy conference that the company was “releasing it to a subset of some of the world’s most important companies and organizations so they can use this to find vulnerabilities.”

Clark cautioned that Mythos would not remain unique for long. “There will be other systems just like this in a few months from other companies, and in a year to a year-and-a-half later, there will be open-weight models from China that have these capabilities,” he said. “So the world is going to have to get ready for more powerful systems that are going to exist within it.”