Anthropic Regains Limited Access to Mythos 5 Model for Select U.S. Users

Anthropic said on June 26 that the government has authorized it to release Claude Mythos 5 to certain U.S. organizations after reversing an earlier suspension tied to security concerns. The company stated the approval covers a small group of cyber defenders and infrastructure providers. It added that it is working to restore access for the approved set as quickly as possible.

Suspension The Trump administration on June 12 issued an export control directive that suspended access to Mythos 5 and Fable 5. Officials cited concerns that a method of jailbreaking Fable 5 had been identified. Anthropic said at the time that jailbreaking refers to exploiting vulnerabilities to bypass built-in safety guardrails.

A letter dated June 26 from the Commerce Secretary informed Anthropic of the change. The letter stated that appropriate safeguards are now in place to permit certain trusted partners to access the model. The same letter noted that Anthropic has committed to working with the government to develop protocols and standards for the AI models.

It did not specify how many companies would receive access or the selection criteria.

“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.”

The U.S. Department of Commerce did not extend the approval to Fable 5.

The designation prevents the company from doing business with the federal government and its contractors. The Pentagon had said in February that it only asked Anthropic to allow use of Claude models for all lawful purposes and had no intention to use AI for mass surveillance or fully autonomous weapons.

OpenAI separately announced on June 26 the preview release of GPT-5.6, which will be limited to a small group of users approved by the Trump administration.