US lifts export ban on Anthropic's most powerful AI models after two-week standoff

The June ban

On June 12, the US Commerce Department abruptly ordered Anthropic to suspend access to its Claude Mythos 5 and Fable 5 AI models, citing national security. Officials feared that the models' ability to identify software vulnerabilities could be exploited by hackers if safety restrictions were bypassed. Anthropic challenged the reasoning but complied, blocking all access, including for its own non-American staff.

Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America's leadership in AI.

A partial thaw and negotiations

After weeks of tense talks, the government partially relented on June 26, allowing Mythos 5 access for a select group of US cyber defenders and infrastructure operators. Foreign partners, including European cybersecurity agencies, remained locked out, stoking concerns about digital sovereignty. The company agreed to proactively detect and address security risks, working with the government on protocols for current and future models.

Full lifting and next steps

On Tuesday, Commerce Secretary Howard Lutnick formally lifted the export controls. In a letter, he said Anthropic had taken steps to address the risks. The company announced that access would be restored starting Wednesday. The move ends a more than two-week standoff that had roiled the AI industry and raised questions about the Trump administration's regulatory posture.

We have received notification that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5. We will begin restoring access tomorrow.

A shift in US AI regulation

The episode marks a sharp turn in US technology policy. Once hostile to AI regulation, the Trump administration has adopted a more hands-on approach, citing cybersecurity threats. Rival OpenAI launched its own GPT-5.6 model last week under a similar arrangement, requiring government vetting of new users. Anthropic's fraught relationship with Washington, the company is suing the Pentagon over a supply-chain risk designation, adds further complexity.