Trump Administration Grants Anthropic Permission to Launch Mythos for Certain US Entities

The US government has relaxed the restrictions on Anthropic’s most advanced AI model, Claude Mythos 5, allowing the company to provide access to over 100 US organizations, including major corporations and government entities.

According to a letter obtained by WIRED from US Commerce Secretary Howard Lutnick to Anthropic’s cofounder and chief compute officer Tom Brown, the government will allow select trusted partners to access Mythos due to the determination that “appropriate safeguards are in place.” Semafor first reported on the letter’s contents.

“Anthropic has collaborated with the U.S. government to mitigate risks associated with the Covered Models. These efforts have led to significant advancements,” Lutnick stated.

However, the government did not approve a wider release of the model and made no mention of the status of Claude Fable 5, the consumer-oriented version of Mythos released by Anthropic with enhanced safeguards. Lutnick highlighted that other requirements specified in his initial directive from June 12 remain unchanged.

“We received notification from the US government that Mythos 5, our most advanced cybersecurity model, can be redeployed to a limited group of cyber defenders and infrastructure providers,” stated Anthropic spokesperson Eduardo Maia Silva in a comment to WIRED. “We are working to set up the authorized providers and restore their access to Mythos 5 as swiftly as possible. We are encouraged by this progress and will continue collaborating with the government to broaden access to Mythos 5 and reintroduce Fable 5 for general use.”

Anthropic is still in talks with the White House about reinstating access to Fable 5, which are expected to continue throughout the weekend, as per a source familiar with the discussions. Both sides are optimistic that the resolution of this situation will contribute to establishing a long-term policy framework for future model launches, the source mentioned.

This partial reinstatement comes about two weeks after the White House issued an export control directive to Anthropic, requiring the company to restrict access to Mythos and Fable 5 for foreign nationals, including those living and working in the U.S. Consequently, Anthropic disabled access to the models entirely. In his recent letter, Lutnick noted that organizations permitted to use Mythos can now allow their foreign national employees access to the model, with Anthropic applying the same for its own foreign national workforce.

The Trump administration expressed concerns regarding Anthropic’s deployment of Mythos after learning that the company had granted access to a South Korean telecommunications firm believed to have connections to China, as previously reported by WIRED. Additionally, Amazon and the National Security Agency separately voiced concerns to the White House about the potential for Fable 5 to be exploited, leading officials to conclude that action was necessary.

Recently, Anthropic dispatched senior members from its cybersecurity and AI safety teams to Washington, DC, to engage with Trump administration officials. Alongside Brown, Anthropic’s public policy chief Sarah Heck has been leading the company’s negotiations with the US Department of Commerce.

Restoring Mythos 5 marks a positive development for both Anthropic and the White House, but the situation has raised broader questions about the trajectory of US AI policy, particularly regarding the extent to which the Trump administration will seek to regulate future model releases. On Friday, OpenAI announced a delay in the launch of its forthcoming GPT 5.6 models following a request from the Trump administration.