Anthropic reinstated access to its AI model Fable after the US government lifted export controls imposed in June. The controls, announced on June 12, barred Anthropic from serving Fable and its sibling model Mythos to non-US citizens, including its own employees, forcing the company to pull the models shortly after Fable's June 9 release, according to combine-lab.github.io.
Following the export restrictions, Anthropic engaged in negotiations that led to the lifting of controls on Fable, allowing general availability to resume. Mythos remains restricted to specific, pre-approved partners. The export controls had made it legally impossible for Anthropic to verify all users and comply with the regulations, prompting the initial withdrawal of the models. The episode generated significant online discussion and media coverage.
The export controls and subsequent lifting highlight the challenges AI companies face regarding national security and international access. Fable, described as a safety-conscious version of Mythos, was intended for broad use but encountered operational hurdles due to regulatory constraints. The situation underscores the complex balance between AI innovation and government oversight, with Anthropic navigating these issues amid heightened scrutiny of AI technologies.
Anthropic's restoration of Fable access came roughly three weeks after the export controls were imposed. Mythos remains limited to select partners, reflecting ongoing regulatory considerations. The initial release date of Fable was June 9, 2026, with the export controls announced three days later, marking a brief but impactful disruption in availability.