Trump Administration Approves Anthropic’s Mythos for Select US Groups

The US government has loosened its grip on Anthropic’s most advanced AI system, Claude Mythos 5, granting the company permission to restore access for more than 100 American organizations, including major corporations and federal agencies. The decision marks a significant shift in the ongoing standoff between the Trump administration and the AI lab over national security concerns tied to cutting-edge models.

Commerce Secretary Howard Lutnick informed Anthropic cofounder and chief compute officer Tom Brown in a letter, obtained by WIRED, that certain trusted partners could once again use Mythos because he had “determined that appropriate safeguards are in place.” The letter, first reported by Semafor, signals a cautious thaw in relations. “Anthropic has worked with the U. S. government to address risks associated with the Covered Models. These efforts have yielded significant progress,” Lutnick wrote.

Still, the administration stopped short of greenlighting a full public rollout. Nothing was said about Claude Fable 5, the consumer-facing version of Mythos that Anthropic released with additional safety measures. Lutnick stressed that other requirements from his initial June 12 directive remain in force, leaving the broader availability of the model in limbo.

“We received notice from the US government that Mythos 5, our strongest cybersecurity model, can be redeployed to a small group of cyber defenders and infrastructure providers,” said Eduardo Maia Silva, an Anthropic spokesperson. “We are working to provision the approved set of providers and restore their access to Mythos 5 as quickly as possible. We are pleased to see this progress and continue to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again.”

Anthropic remains in active talks with the White House about restoring Fable 5, with discussions expected to continue through the weekend, according to a person familiar with the matter. Both sides are reportedly hopeful that resolving this episode will help shape a lasting policy framework for future model releases.

The partial reinstatement comes roughly two weeks after the White House issued an export control directive requiring Anthropic to restrict foreign nationals,including those living and working in the US,from accessing Mythos and Fable 5. In response, the company disabled access to both models entirely. Lutnick’s latest letter now permits approved organizations to allow their foreign national employees to use Mythos, and Anthropic can do the same for its own foreign-born staff.

The Trump administration’s concerns first emerged after it learned that Anthropic had granted access to a South Korean telecommunications firm believed to have ties to China, as WIRED previously reported. Separately, Amazon and the National Security Agency flagged worries that Fable 5 could be jailbroken. These converging alarms convinced officials that immediate action was necessary.

In recent weeks, Anthropic dispatched senior members from its cybersecurity and AI safety teams to Washington, DC, for meetings with Trump administration officials. Alongside Brown, Anthropic’s public policy chief Sarah Heck has been leading discussions with the US Department of Commerce.

While getting Mythos 5 back online represents a promising step for both Anthropic and the White House, the broader saga has raised fundamental questions about the trajectory of US AI policy,especially how far the Trump administration will go in controlling future model releases. On Friday, OpenAI announced it was delaying the launch of its upcoming GPT 5.6 models in response to a similar request from the administration, underscoring the growing tension between innovation and government oversight.