Trump Admin Permits Anthropic's Mythos 5 for US Organizations

The Trump administration has reversed course on the most advanced AI model from Anthropic — Claude Mythos 5 — authorizing deployment to over 100 US organizations, including large corporations and government agencies. Commerce Secretary Howard Lutnick signed a letter stating that appropriate safeguards are now in place after Anthropic worked with the government to address risks. The reversal comes two weeks after a June 12 export control directive forced Anthropic to limit foreign national access to both Mythos and the consumer variant Fable 5, prompting the company to disable the models entirely.

Under the new guidance, approved organizations and Anthropic itself can allow foreign national employees to use Mythos 5. The company is now restoring access to cybersecurity defenders and infrastructure providers. However, the White House stopped short of permitting a broader public release, and Fable 5 remains blocked. Anthropic is in active weekend talks with the administration to resolve Fable 5’s status, with both sides hoping the episode will inform a lasting AI policy framework.

The controversy triggered after Anthropic gave access to a South Korean telecom with alleged China ties, and separate Amazon/NSA concerns about jailbreak risks. The saga signals a new era of government control over frontier AI releases. OpenAI recently delayed its GPT-5.6 model in response to White House requests, and former adviser Dean Ball noted that frontier developers now “need an explicit green light from the government.” Anthropic’s battles have been costly — it sued the administration earlier this year over supply chain designations tied to military contractors.