The United States government has lifted export restrictions that had frozen access to Anthropic’s most advanced artificial intelligence models, the company confirmed, clearing the way for the systems to return to users after weeks in limbo.
Anthropic said it had been notified that the Department of Commerce had lifted the controls on its Fable 5 and Mythos 5 models and that it would begin restoring access the following day. The announcement came shortly after Commerce Secretary Howard Lutnick said on social media that his department had worked closely with Anthropic over the preceding two weeks to review and approve Fable 5, describing the move as a step to strengthen America’s leadership in AI.
Fable is a version of Anthropic’s flagship Mythos model fitted with additional safeguards intended to make it suitable for public release. The export ban had forced Anthropic to suspend access for all foreign nationals, including its own employees, after a trusted partner reportedly identified by CNN as Amazon discovered a way to bypass Fable’s safety guardrails. Anthropic described the workaround at the time as simple and noted that other publicly available models had similar vulnerabilities, while a company source said a new safeguard had since been implemented to close the gap.
Commerce officials had allowed Anthropic to release Mythos to a small group of government-approved entities the previous week, citing concerns that the model’s capabilities could be exploited to identify cybersecurity vulnerabilities at an unusually rapid pace. Anthropic has limited that release to a select group of partners focused on securing critical software systems.
The episode is part of a broader pattern of regulatory caution around powerful AI systems. The White House has also asked OpenAI to limit distribution of its upcoming GPT 5.6 model to vetted government partners, a request OpenAI said it hoped would not become a long-term standard for AI releases.
Source: edition.cnn.com