White House Chief of Staff meets with Anthropic CEO about new Mythos AI model

WASHINGTON (AP) — White House chief of staff Susie Wiles plans to poll Anthropic CEO Dario Amodei about the artificial intelligence company Mythos’ new model, which has caught the attention of the federal government on how to transform national security and the economy.

He watches: Anthropic’s new powerful AI model raises concerns about the risks of advanced technology

A White House official, who requested anonymity to discuss the meeting scheduled for Friday, said the administration is communicating with advanced artificial intelligence laboratories about their models and software security. The official stressed that any new technology that the federal government may use will require a technical period of evaluation.

The meeting comes after escalating tensions between the Trump administration and safety-conscious Anthropic, which has sought to place barriers to the development of artificial intelligence to reduce any potential risks and maximize its economic and national security benefits for the United States.

President Donald Trump tried to ban all federal agencies from using Anthropic’s chatbot, Cloud, over the company’s contract dispute with the Pentagon, with Trump saying in a social media post in February that the administration would “never do business with them again!”

Defense Secretary Pete Hegseth also sought to have Anthropic declared a supply chain risk, an unprecedented move against a US company that Anthropic has challenged in two federal courts. The company said it wanted guarantees that the Pentagon would not use its technology in fully autonomous weapons and to monitor Americans. Hegseth said the company should allow any uses the Pentagon deems legal.

US District Judge Rita Lynn issued a ruling in March blocking implementation of Trump’s social media directive ordering all federal agencies to stop using humanitarian products.

He watches: Why is the Trump administration clashing with the artificial intelligence company Anthropic?

Anthropic declined to discuss the meeting in advance.

San Francisco-based Anthropic said the new Mythos model it announced on April 7 is so “strikingly capable” that it limits its use to select customers because of its ability to outpace human cybersecurity experts in finding and exploiting computer vulnerabilities.

While some industry experts questioned whether Anthropic’s claims of extremely powerful AI technology were a marketing ploy, some of the company’s harshest critics have suggested that Mythos may represent a further advance in AI.

David Sachs, an influential anthropology critic who was the White House’s AI and cryptocurrency official, said people should “take this seriously.”

“Anytime Anthropology scares people, you have to ask: Is this a tactic? Is this part of their Chicken Little routine? Or is it real?” Sachs said on the “All-In” show he hosts with other tech investors. “As far as the internet goes, I’ll give them credit in this case and say this is more about the real thing.”

“It stands to reason that as programming models become more capable, they become more able to detect bugs. That means they’re more able to find vulnerabilities. That means they’re more able to string multiple vulnerabilities together and create vulnerabilities,” Sachs said.

The potential benefits of this model, as well as its risks, have also attracted attention outside the United States

The UK’s AI Security Institute said it had evaluated the new model and found it to be a “step forward” compared to previous models, which were already improving rapidly.

“Mythos Preview can exploit systems with weak security posture, and it is likely that more models will be developed with these capabilities,” the institute said in a report.

Anthropic is also in talks with the EU about its own AI models, including advanced models that have not yet been released in Europe, European Commission spokesman Thomas Regnier said on Friday.

Axios first reported the scheduled meeting between Wells and Amodei.

When it announced Mythos, Anthropic said it was also forming an initiative called Project Glasswing, bringing together tech giants like Amazon, Apple, Google and Microsoft, along with other companies like JPMorgan Chase, in hopes of securing the world’s critical software from the “severe” ramifications the new model could have on public safety, national security and the economy.

“We’re releasing it to a subset of the most important companies and organizations in the world so they can use it to find vulnerabilities,” Jack Clark, co-founder and head of policy at Anthropic, said at the Semaphore Global Economics conference this week.

Clark added that Mythos, while ahead of the curve, is not a “special model.”

“There will be other systems like this within a few months from other companies, and within a year to a year and a half, there will be open-weight models from China with these capabilities,” he said. “So the world will have to prepare for more powerful systems that will exist within it.”

O’Brien reported from Providence. RI AP business reporter Kelvin Chan contributed to this report from London.