Expectations are rising that the Donald Trump administration in the U.S. will soon lift restrictions on Anthropic’s cutting-edge artificial intelligence (AI) model, Claude Fable 5, potentially as early as next week. Meanwhile, Claude Mythos 5, which was also subject to access limitations, may now be available again to select verified users.
On June 27, Axios reported, citing sources close to the matter, that the Trump administration is poised to ease restrictions on Fable 5 usage in the near future.
Another insider suggested that discussions on this topic are set to continue over the weekend, with rumors circulating about Anthropic’s imminent restoration of Fable 5 access.
However, the final green light for user access hinges on approval from the U.S. Department of Defense and the National Security Agency (NSA). Axios cautioned that without these approvals secured, it’s premature to confirm a definitive timeline for the model’s return.
On June 26, the U.S. Department of Commerce granted permission for Mythos 5 to be made available again, albeit exclusively to a select group of government-approved users.
Foreign media outlets report that Anthropic released a statement saying that it has been collaborating closely with the U.S. government to reinstate access to Mythos 5 and Fable 5. The company added that it has received notification that it can redistribute Mythos 5, the most robust cybersecurity model, to various U.S. agencies responsible for operating and safeguarding critical infrastructure.
Earlier, on June 12, the U.S. government had issued export control guidelines that completely halted access to Mythos 5 and Fable 5, citing national security concerns.
U.S. Secretary of Commerce Howard Lutnick sent a letter to Anthropic, mandating that individual permission must be sought to provide or export Mythos 5 and Fable 5 to foreign entities.
In response, Anthropic opted to entirely suspend access to both models, citing the impracticality of verifying each user’s nationality. This move has impeded the use of these models by the South Korean government and several companies that had joined Project Glasswing, which offered access to Mythos.
Mythos 5 is a model tailored for cybersecurity, incorporating safeguards to prevent its misuse in hacking, cyberattacks, or biological terrorism. Fable 5 is renowned for its advanced reasoning capabilities and swift coding performance. Anthropic has previously asserted that Fable 5 is the most powerful model among publicly available AI systems.