After weeks of negotiations with the Trump administration, US artificial intelligence company Anthropic has finally confirmed that its consumer-facing Claude Fable 5 model will be back online. The company said on social platform
Anthropic said in a statement that it has received a notification from the U.S. Department of Commerce confirming that export restrictions on Claude Fable 5 and Mythos 5 have been lifted. The company will resume access to related models starting from the next day and will announce more operational arrangements later. The company also expressed its gratitude to users who have been patient during the shutdown and to all parties who assisted in the redeployment of models.
In early June this year, Anthropic took the initiative to offline Fable 5 after receiving an ultimatum issued by the Trump administration on Friday night. This model is aimed at consumers and is based on the same underlying technology as Mythos 5, but has stricter security protections and usage restrictions. Because the government is concerned that related technologies may be "jailbroken" to bypass security measures, the Trump administration issued an export control order to Anthropic, prohibiting any "foreign persons" (including non-U.S. employees of corporate client companies, and even a large number of foreign employees within Anthropic companies) from using Mythos 5 or Fable 5. This ban directly blocks international access to the two core models that Anthropic has highlighted in the past week.
Before Fable 5 was allowed to return, the Trump administration had taken the lead in agreeing to allow Mythos 5 to return to service with conditions. According to the approval plan, the first phase of Mythos 5 will only be open to a "list of pre-vetted organizations." Organizations on the list can regain access to the model, and their non-U.S. employees and Anthropic's foreign employees are also allowed to use the model. The timing of this arrangement coincides with the release of GPT-5.6 by rival company OpenAI, and it is also carried out under the rules set by the Trump administration: the new model adopts a batch launch strategy, and is initially only open to testing and use by organizations and relevant departments pre-approved by the government.
The timing at which Anthropic encountered export controls is considered quite sensitive. On the one hand, the company is preparing to advance its initial public offering (IPO); on the other hand, it has continued to have disputes with the US government over the past few months over "supply chain risk identification." Previously, the U.S. Department of Defense and relevant agencies made regulatory rulings related to supply chain risks against Anthropic, and Anthropic filed a lawsuit and applied for an injunction through legal channels. The conflict between the two parties continues to this day. Against this background, the export restrictions and negotiations surrounding Fable 5 and Mythos 5 are regarded by the outside world as the latest confrontation between Anthropic and the White House on artificial intelligence regulation and national security issues.
As export controls are lifted, Anthropic's consumer-oriented Fable 5 and enterprise-level Mythos 5 are gradually returning to the market. Under the review framework of the Trump administration, the redeployment of these two models still needs to comply with specific access rules and security requirements, but their re-launch will undoubtedly bring new variables to Anthropic's business expansion, capital market expectations, and model options for global users.
Related articles:
Sources say Claude Fable 5 is expected to resume visits this week Anthropic dispatches commissioners to discuss with US government
US to allow Anthropic to restore access to Fable5 model