White House chief of staff to meet with Anthropic CEO about dangerous new Mythos model, official says

A White House official, who asked not to be identified to discuss Friday’s planned meeting, said the administration is working with advanced AI labs about their models and software security. The official emphasized that any new technology that may be used by the federal government will require a technical period for evaluation.

The meeting comes after tensions have heated up between the Trump administration and safety-minded Anthropic, which has intends to install guardrails of AI development to minimize any potential risks and maximize economic and national security benefits for the US

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

Defense Secretary Pete Hegseth also sought to declare Anthropic a supply chain hazard, an unprecedented move against a US company that Anthropic has challenged in two federal courts. The company said it wanted assurances that the Pentagon would not use its technology fully autonomous weapons and American surveillance. Hegseth said the company must allow any use the Pentagon deems lawful.

US District Judge Rita Lin issued a ruling in March which blocked the implementation of Trump’s social media directive ordering all federal agencies to stop using Anthropic products.

Anthropic declined to talk about the meeting in advance.

San Francisco-based Anthropic says the new Mythos model it announced on April 7 is “uniquely capable” of limiting its use to select customers because of its ability to outperform human cybersecurity experts in finding and exploiting computer vulnerabilities.

And while some industry experts question whether Anthropic’s claims of more powerful AI technology are a marketing ploy, even some of the company’s sharpest critics suggest that Mythos may represent a further advance in AI.

An influential critic of Anthropic, David Sacks, who is AI and crypto czar in the White House, said that people should “take this seriously.”

“Anytime Anthropic scares people, you have to ask, ‘Is this a ploy? Is this part of their Chicken Little routine? Or is it real?'” Sacks said on the “All-In” podcast he co-hosts with other tech investors. “In cyber, I’ll definitely give them credit in this case and say it’s more than the real part.”

Sacks said, “It makes sense that as coding models become more and more capable, they’re better able to find bugs. That means they’re better able to find vulnerabilities. That means they’re better able to combine multiple vulnerabilities and create an exploit.”

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

The United Kingdom’s AI Security Institute said it had tested the new model and found it to be a “step up” from previous models, which were rapidly improving.

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

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

Axios first reported the scheduled meeting between Wiles and Amodei.

When it announced Mythos, Anthropic said it was also forming an initiative called Project Glasswing, bringing together tech giants such as Amazon, Apple, Google and Microsoftalong with other companies like JPMorgan Chasein hopes of securing the world’s critical software from the “severe” fallout that the new model could bring to public safety, national security and the economy.

“We’re releasing it to a subset of some of the most important companies and organizations in the world so they can use it to find vulnerabilities,” said Anthropic co-founder and chief policy officer, Jack Clark, at the Semafor World Economy conference this week.

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

“There will be other systems 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 with these capabilities,” he said. So the world needs to prepare for more powerful systems to exist within it. ”

___

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