The US government has eased restrictions on Anthropic’s most advanced AI model, Claude Mythos 5, permitting the AI lab to grant access to more than 100 US organizations, including large corporations and government agencies, according to a letter obtained by WIRED.
The partial reinstatement came roughly two weeks after the White House sent an export control directive requiring Anthropic to limit foreign national access to both Claude Mythos 5 and the consumer-facing Claude Fable 5, including people working and living in the United States.
Letter from Commerce Secretary
In a letter sent to Anthropic’s cofounder and chief compute officer Tom Brown, Commerce Secretary Howard Lutnick stated he had “determined that appropriate safeguards are in place” and would permit certain trusted partners to regain access. Lutnick wrote:
“Anthropic has worked with the U.S. government to address risks associated with the Covered Models. These efforts have yielded significant progress.”
However, the White House stopped short of permitting a broader rollout and said nothing about the fate of Claude Fable 5. Lutnick noted that all other requirements in the initial directive sent on June 12 remain in effect.
Approved Access and Restrictions
Approved organizations may now extend access to their foreign national employees and Anthropic’s foreign national employees, reversing part of the earlier restriction. Anthropic spokesperson Eduardo Maia Silva confirmed:
“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. We are working to provision the approved set of providers and restore their access to Mythos 5 as quickly as possible.”
Background of Restrictions
The Trump administration developed concerns after learning Anthropic had shared access to Mythos with a South Korean telecommunications firm believed to have ties to China. Additionally, Amazon and the NSA raised concerns that Claude Fable 5 could be jailbroken, convincing the administration to act.
In recent weeks, Anthropic sent senior members from its cybersecurity and AI safety teams to Washington, DC to meet with officials. In addition to Tom Brown, public policy chief Sarah Heck led discussions with the US Department of Commerce.
Industry Implications
On Friday, OpenAI announced it was delaying the release of its upcoming GPT 5.6 models in response to a request from the Trump administration. Dean Ball, head of the strategic futures team at OpenAI and a former White House AI adviser, said in a June 16 blog post that Anthropic’s latest spat with the Trump administration has shown frontier AI model developers they “need an explicit green light from the government now.”
Anthropic’s Legal and Business Challenges
Anthropic sued the Trump administration earlier this year over a supply chain risk designation it received after trying to set redlines around how military contractors could use its AI models. WIRED reported that some Anthropic investors worked to determine what the White House’s restrictions on Claude Fable 5 meant for Anthropic’s corporate future. Getting Mythos 5 back online is a promising step, but the saga has raised broader questions about US AI policy and the extent of government control over future model releases.
Sources:
