Signal Frontier model export control suspension and reinstatement (Fable 5 / Mythos 5)

Summary

On June 9, 2026, Anthropic released Claude Fable 5, the first publicly available model in its Mythos-class tier; days later, on the evening of June 12, the US Department of Commerce directed Anthropic to suspend all access to Fable 5 and Mythos 5 for any foreign national, including the company's own non-citizen employees, reportedly giving Anthropic 90 minutes to comply. Because Anthropic could not reliably identify foreign nationals across its global user base in real time, it implemented a universal shutdown, taking both models offline for all customers across every integration platform simultaneously. This marked the first publicly documented exercise of export control authority over a commercially deployed frontier AI model; Commerce lifted the controls on June 30, and the models returned on July 1. When Anthropic released Mythos 5 in April 2026, it deemed the model too dangerous for public release and initially allowed only a few dozen trusted cybersecurity vendors, major technology companies and the US government to test it under a project called Glasswing. OpenAI separately delayed the public release of GPT-5.6 at government request, and the episode spurred a legislative response, Representative Lieu's Kill Switch Act, which cites the Fable/Mythos episode as a motivating incident. Together these events form a pattern in which frontier model access is treated as national-security leverage, extending a roughly four-year US export-control ascent from chips and manufacturing equipment to compute, cloud access and now the models themselves.