U.S. Commerce Department lifts export block on Anthropic's Claude Mythos 5 model

The U.S. Commerce Department said on Friday that it has removed the export restriction on Anthropic’s Claude Mythos 5 artificial-intelligence model, allowing the company to share the system with more than 100 U.S. institutions, including major corporations and government agencies. In a letter to Anthropic’s chief compute officer Tom Brown, Commerce Secretary Howard Lutnick wrote that “appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model,” noting “significant progress” in daily talks since the block was imposed two weeks earlier.

The earlier restriction had halted distribution of Mythos and its smaller counterpart, Fable 5, after concerns from Amazon and other firms that the models could be modified for malicious use. The current letter does not mention Fable 5, but sources close to the negotiations said a release may be considered in the future.

Anthropic has agreed to cooperate with the U.S. government on protocols, standards and future releases of its models, according to the letter. The new arrangement also eliminates the need for a license to export, re-export, or transfer Claude Mythos 5 to entities listed in Annex A of the letter and to foreign-national employees of those entities or Anthropic.

The decision coincides with OpenAI’s limited launch of its GPT-5.6 model to a short list of government-approved partners. Commerce Department spokesperson Benno Kass said the agency acted quickly to address security concerns while maintaining U.S. leadership in AI development.

European officials and other U.S. allies have expressed concerns about reliance on Washington’s decisions for access to advanced AI tools. Earlier reports indicated that the U.S. had raised issues about Mythos being shared with partners perceived to have close ties to China, including a South Korean telecommunications firm.