The US has removed export restrictions on Anthropic's Fable 5, allowing the model to be reintroduced.

      The US Commerce Department has removed the export controls it had imposed on Anthropic's most advanced models, concluding a roughly three-week suspension that took the company's Claude Fable 5 offline. Reuters first reported this change on Tuesday, 30 June, citing an insider familiar with the decision. Shortly thereafter, Anthropic officially confirmed it in a news post titled “Redeploying Claude Fable 5”.

      The controls affected both Fable 5, which is the public model, and Mythos 5, the more powerful cybersecurity framework that underlies it. Both models were cleared simultaneously. Commerce Secretary Howard Lutnick announced that an export license would no longer be necessary. In return, Anthropic committed to proactively identifying and managing security risks, assisting in the development of standards for future models, and reporting any malicious activities to the government.

      These restrictions had been enforced around 12 June, and were reversed about two and a half weeks later. The initial order mandated the shutdown of both Fable 5 and Mythos 5 almost immediately. The trigger for the actions was a jailbreak, where researchers, reportedly from Amazon, discovered a method to circumvent Fable 5’s safety measures and access the underlying capabilities of the Mythos model.

      Throughout the incident, Anthropic maintained that the vulnerability was not serious enough to require a complete shutdown, while regulators initially took a stricter stance. The differentiation between the two systems is significant: Mythos 5 is the robust cybersecurity model, while Fable 5 serves as the consumer-facing version built on it, which explains why a jailbreak of one raised concerns about the other.

      The initial restrictions may have been motivated by fears that a group connected to China could exploit the vulnerability in Mythos. This concern is based on anonymous reports and should be regarded as speculative rather than confirmed.

      The government’s position appears to have evolved in the following weeks, with several media outlets suggesting that the reversal was partly a response to industry criticism and concerns that international competitors could gain ground while a leading US model remained inactive, although this interpretation is analytical rather than an official statement.

      When will the models return? Anthropic announced that Fable 5 would be reinstated globally on 1 July across its products, including the Claude platform, Claude.ai, Claude Code, and Claude Cowork. The rollout will be gradual, with usage capped at 50% of the normal weekly limits until 7 July, after which full access will resume.

      The three-week episode has been notably public for an export-control situation, which typically unfolds quietly through the Bureau of Industry and Security. In this instance, the model was specifically mentioned, and the company provided real-time updates about its suspension and subsequent revival.

      This situation is part of a broader tightening of US policy regarding the global distribution of advanced AI technologies. Over the past year, Washington has been addressing loopholes that previously allowed cutting-edge chips and models to reach Chinese companies through foreign subsidiaries.

      For Anthropic, the immediate implications were as commercial as they were regulatory. A halted flagship model meant potential customers would seek alternatives, and the competitive disadvantage grew day by day. The company spent a significant amount of June engaging with officials in Washington to advocate its position as the freeze continued. The conditions set by Lutnick for lifting the suspension seemed to be the price for regaining access.

      Anthropic had previously been authorized to reactivate Mythos 5 for a select group of trusted cybersecurity partners before this week, while Fable 5 remained restricted. Tuesday’s decision effectively closed that remaining gap.

      The confirmed facts are narrow and straightforward: Commerce lifted the controls, Lutnick established conditions, Anthropic accepted them, and access is being restored. However, the political implications surrounding the reversal, including the specific role of the alleged China connection and the impact of competitive pressures, remain less defined. On these matters, the information is limited, and a more comprehensive account will likely emerge in the coming weeks.