The public will once again have access to Anthropic’s powerful Fable 5 artificial intelligence system, the company confirmed on Tuesday. This development comes just weeks after the U.S. Commerce Department mandated that Anthropic disable the model due to concerns regarding potential security risks. According to the company, access is scheduled to be restored for customers beginning Wednesday.
Fable 5, which operates as part of the broader Claude AI family, was pulled offline on June 12 alongside its counterpart, the Mythos 5 model. At the time, senior administration officials stated that these systems posed significant cybersecurity threats and alleged that Anthropic’s leadership had failed to adequately address these concerns. This temporary suspension marked a notable shift in the government’s regulatory posture, indicating that AI technologies have reached a level of power necessitating increased federal oversight.
Commerce Secretary Howard Lutnick communicated the decision to lift export control restrictions in a letter to Anthropic co-founder Tom Brown on Tuesday. The decision followed a period of intense collaboration between the company and government officials to mitigate the risks associated with the software. In the letter, Lutnick noted that Anthropic has committed to ongoing collaboration regarding safety protocols, future releases, and immediate notification of any detected malicious activity.
On the social media platform X, Secretary Lutnick confirmed the decision, stating that the government worked closely with Anthropic over the past two weeks to analyze the model and ensure alignment with national standards. Previously, on Friday, Lutnick had granted a limited redeployment of Mythos 5, allowing access for approximately 100 trusted organizations involved in cybersecurity and infrastructure protection. Anthropic stated it is working to restore Mythos 5 access to a broader range of users, though currently, only a subset of U.S.-based entities retain that access.
Anthropic launched Fable 5 and Mythos 5 on June 9, representing the company’s most advanced systems to date. While Fable 5 originally included built-in guardrails to prevent use in harmful cybersecurity or biological tasks, officials feared these protections could be circumvented. In response, Anthropic dispatched top scientists to Washington to negotiate a resolution. The company has since implemented a new method to target and block potential workarounds, a solution that the Commerce Department’s Center for AI Standards and Innovation has tested and endorsed.
During the period the models were offline, industry experts expressed concerns that such ad-hoc bans could impact American leadership in the global AI race. OpenAI CEO Sam Altman, a competitor to Anthropic, criticized the government’s push for phased, government-approved releases. The regulation of these models follows an executive order signed by President Donald Trump in early June, which seeks to establish voluntary mechanisms for vetting advanced AI systems before they are released to the public.
