The United States government has officially authorized Anthropic to provide access to its advanced artificial intelligence model, Mythos 5, to a select group of American cybersecurity organizations. This development was confirmed by the company on Friday, marking a shift in its operational capabilities regarding its most powerful security-focused software.
“Today, the government notified us that Mythos 5, our strongest cybersecurity model, can be redeployed to a set of US organizations that operate and defend critical infrastructure,” Anthropic stated in its announcement. The company added that it is working to restore access for these entities rapidly, while continuing discussions with federal authorities to broaden the availability of both Mythos 5 and the Fable 5 model for general public use in the future.
This policy update arrives just two weeks after the Trump administration enforced a block on the use of Mythos 5. The company’s relationship with the current administration has been marked by tension for several months. Previously, on June 12, Anthropic was forced to abruptly withdraw both Mythos 5 and Fable 5 from all users following a government directive that restricted their deployment.
The Trump administration had justified the initial suspension by raising concerns that these sophisticated AI systems could potentially be exploited by military intelligence agencies in countries like Russia or China. These actions were part of a broader, more aggressive oversight strategy regarding the release of frontier models developed by both Anthropic and its competitor, OpenAI. The administration has faced criticism for these strict control measures, with some accusing the government of regulatory overreach.
Details regarding the selection process remain opaque; the administration has not publicly clarified which specific companies will receive access or the criteria used to vet them. This lack of transparency drew criticism from OpenAI CEO Sam Altman, who noted on social media that while extensive safety testing is a valid approach, he expressed discomfort with the government acting as a gatekeeper to choose specific customers for top-tier models.
According to a source familiar with the directive who spoke on the condition of anonymity, more than 100 institutions and companies—including various Fortune 500 firms—are now slated to gain access to the Mythos 5 model.
