The White House has lifted restrictions on Anthropic’s AI model, Claude Mythos 5, allowing access to over 100 U.S. organizations, including major corporations and government entities. In a recent letter, U.S. Commerce Secretary Howard Lutnick stated that trusted partners could now use Mythos, indicating that sufficient safeguards are in place.
Anthropic’s spokesperson confirmed the model’s restoration to a select group of cyber defenders and infrastructure providers. However, the fate of its consumer-facing model, Fable 5, remains unclear as discussions between Anthropic and the government are ongoing. Lutnick emphasized that the original constraints from prior directives continue to apply.
This decision follows the administration’s demand to restrict access to Mythos and Fable 5, leading to Anthropic temporarily disabling the models. Concerns had arisen from the administration regarding a South Korean telecommunications firm’s ties to China and from Amazon’s and the NSA’s worries over the potential for the models to be manipulated.
In light of these challenges, Anthropic engaged with White House officials, sending key team members for discussions to navigate regulatory obstacles. The careful reinstatement of Mythos is considered an important step for both Anthropic and the government but has sparked broader discussions about the future of AI policy in the U.S.
Recent tensions resulted in OpenAI delaying the release of its GPT 5.6 models due to similar requests from the administration. The ongoing developments suggest that AI model developers may now require explicit government approval before launching new models.
