Anthropic will begin restoring global access to its Claude Fable 5 and Claude Mythos 5 AI models after the U.S. Department of Commerce lifted export controls, the company said.
The San Francisco-based AI lab said on X that it had received notice that the restrictions on the two models had been withdrawn.
Anthropic said it would begin restoring access on Wednesday. The U.S. government had forced Anthropic to cut off access to the models on June 12. It did so after finding vulnerabilities in safeguards meant to prevent misuse.
Politico reported that U.S. Commerce Secretary Howard Lutnick told Anthropic the Trump administration had withdrawn the previous restrictions. Lutnick’s letter said that Anthropic had worked with the U.S. government to address risks associated with the covered models.
The reversal came four days after Anthropic received authorisation to allow a small group of American cybersecurity firms to access Mythos 5. OpenAI has also restricted the release of GPT-5.6 to approved partners. This was done after Washington’s request.
OpenAI Chief Executive Officer Sam Altman said on X that the process was not the one his company considered optimal.
Read: Anthropic AI Shutdown Raises Investor Risk Concerns
Anthropic said in a Tuesday blog post that it wanted a standard framework to assess and respond to critical vulnerabilities in advanced AI models. The company said it would work with Amazon, Microsoft, Google and others on that effort.
Central Intelligence Agency Director John Ratcliffe compared frontier AI model capabilities to “digital nuclear weapons” during a speech at the AWS summit in Washington.
