Anthropic Relaunches Claude Fable 5 Globally After Export Restrictions Lifted

Anthropic resumed global availability of its leading artificial intelligence model "Claude Fable 5" on July 1, 2026. This followed the U.S. Department of Commerce's revocation of an emergency export control order on the model the previous evening. The company's official X account announced the resumption at 3:31 p.m. Eastern Time.

This export control traced back approximately three weeks. On June 12, 2026, the U.S. Department of Commerce issued an emergency export control order against both Fable 5 and its sister model "Claude Mythos 5," which specializes in cybersecurity applications. In response, Anthropic halted global access to both Fable 5 and Mythos 5, rendering both models unusable just days after their release. Because the restrictions were imposed through an extraordinary emergency order procedure, criticism quickly arose among artificial intelligence and cybersecurity policy experts.

The scope of Fable 5's resumption extends across the company's major services, including Claude Platform, Claude.ai, Claude Code, and Claude Cowork. Regarding access through enterprise cloud, the company stated it would proceed with reactivation "as expeditiously as possible" on three cloud platforms: Amazon Web Services, Google Cloud, and Microsoft Foundry. However, at the time of this article's writing, restoration on these platforms had not been confirmed.

Mythos 5, by contrast, faces changing regulatory treatment but has not yet achieved public release. According to a letter posted on social media by Commerce Secretary Howard Lutnick, licenses are no longer required for exports, re-exports, or domestic transfers of both Fable and Mythos models. However, Anthropic's announcement of redeployment on its website states that for Mythos 5, access was restored only to "certain U.S. organizations" that received government approval on June 26. The company stated it is continuing to coordinate with the government regarding expansion to additional domestic and international partners through an opt-in cybersecurity verification program called "Project Glasswing."

The background of the regulatory rollback reflected perspectives from both government and industry. Commerce Secretary Lutnick stated that the government and Anthropic worked closely on "detailed analysis and approval" of Fable 5, while White House Chief of Staff Susie Wiles stated on X that "the United States is the leader in the artificial intelligence race, and rapidly and safely distributing the best technology is a shared priority." Regarding the initial control order, former Facebook security executive Alex Stamos characterized the restriction on Fable 5 as "serious self-harm for the United States," warning of the risk of companies turning to Chinese-made models, and such resistance from industry and policy experts continued.

This sequence of events may be viewed as a case that reexamines the proper role of policy in export control of artificial intelligence models. The market disruption caused by short-term restrictions through emergency order and the subsequent rapid policy reversal again demonstrated the difficulty of regulating powerful artificial intelligence models. For Mythos 5, while the legal export restrictions have been lifted, the framework of government-supervised access remains in place, and attention focuses on which organizations and regions will gain access going forward.