SAN FRANCISCO — Anthropic has restored global access to its Claude Fable 5 artificial intelligence model after the U.S. Department of Commerce lifted export controls that had forced the company to suspend the service worldwide in June, ending a disruption that highlighted growing government scrutiny of advanced AI systems.
The company said Fable 5 became available again beginning July 1 across Claude.ai, the Claude Platform, Claude Code and Claude Cowork. Anthropic said it is also working to restore availability through cloud partners including Amazon Web Services, Google Cloud and Microsoft Foundry.
Anthropic announced that the Commerce Department had lifted export controls on both Claude Fable 5 and Claude Mythos 5 after restrictions imposed on June 12 required the company to block access by foreign nationals. Because the order took immediate effect and the company said it had no reliable method to verify users' nationality in real time, it temporarily suspended access for all users rather than only affected groups.
Under the phased restoration, Anthropic said subscribers to Pro, Max, Team and selected Enterprise plans will be able to use Fable 5 for up to half of their weekly usage limits through July 7 before the model transitions to a usage-credit system. Mythos 5, the company's more restricted model, has resumed service only for a limited group of approved U.S. organizations following government authorization granted in late June.
The export controls followed U.S. government concerns about the cybersecurity capabilities of the company's newest AI models after researchers identified a method that could bypass some of Fable 5's safeguards and generate potentially harmful software vulnerability information. Anthropic said it has since implemented additional protections, including a classifier designed to detect and block the identified technique while allowing most legitimate uses to continue.
The company also said it is expanding cooperation with U.S. authorities on AI safety testing, threat reporting and evaluation standards for advanced models. According to Anthropic, the latest safeguards were reviewed before the export controls were lifted.
The episode marked one of the first instances in which U.S. export control powers temporarily halted public access to a commercially available frontier AI model, underscoring increasing regulatory attention to systems with advanced cybersecurity capabilities. Anthropic said it will continue restoring access through additional platforms while working with government agencies on future deployment and safety measures.
