The Trump Administration Lets Anthropic Release the Myth to Choose US Organizations
The US government dropped restrictions on Anthropic’s most advanced AI model, Claude Mythos 5, allowing the company to provide access to more than 100 US organizations, including large corporations and government agencies.
In a letter sent by Anthropic founder and chief computer officer Tom Brown obtained by WIRED, US Commerce Secretary Howard Lutnick told the AI lab that it will allow certain trusted partners access to Mythos because it “has determined that there are appropriate safeguards in place.” Semafor first reported the existence of this book.
“Anthropic has worked with the US government to address the risks associated with Integrated Models. These efforts have brought significant progress,” Lutnick wrote.
However, the government stopped short of allowing a wide release of the model, and said nothing about the fate of Claude Fable 5, the consumer-facing version of Mythos that Anthropic released with significant additional protections. Lutnick noted in his letter that some of the requirements outlined in the original order he sent on June 12 still apply.
“We received notification from the US government that Mythos 5, our most powerful cybersecurity model, can be deployed again to a small group of cyber defenders and infrastructure providers,” Anthropic spokesman Eduardo Maia Silva said in a statement to WIRED. “We are working to provide an approved set of providers and restore their access to Mythos 5 as soon 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 is still in talks with the White House about restoring access to Fable 5, and those are expected to continue over the weekend, according to a person familiar with the matter. Both sides hope that the resolution of the incident will help inform a permanent policy framework for future model releases, the person said.
The partial refund comes nearly two weeks after the White House sent an export control order to Anthropic that required the company to restrict foreign access to Mythos and Fable 5, including people who work and live in the United States. In response, Anthropic disabled access to the models entirely. In his latest letter, Lutnick wrote that organizations authorized to use Mythos can now allow their external employees to access the model, and Anthropic can do the same for its external employees.
The Trump administration is concerned about Anthropic’s release of Mythos after it discovered the company had provided access to a South Korean telecommunications company believed to have ties to China, WIRED previously reported. Amazon and the National Security Agency have also separately raised concerns in the White House that Fable 5 may be arrested, and the confluence of events convinces officials that they need to take action.
In recent weeks, Anthropic sent senior members from its cybersecurity and AI teams to Washington, DC to meet with Trump administration officials. Along with Brown, Anthropic’s chief public policy officer Sarah Heck has been leading the company’s negotiations with the US Department of Commerce.
Bringing Mythos 5 online marks a promising step forward for Anthropic and the White House, but the saga has raised broader questions about the direction of US AI policy, particularly the extent to which the Trump administration will try to control future model releases. On Friday, OpenAI announced that it is delaying the release of its upcoming versions of GPT 5.6 in response to a request from the Trump administration.
