Mythos Comes Back But Not for Everyone

On Friday afternoon, the U.S. government approved a limited reintroduction of Anthropic’s Mythos model, while simultaneously pressuring OpenAI to restrict access to its newly announced GPT-5.6 family. The result is a de facto licensing regime for frontier AI, one that has no legislative mandate, no public transparency, and no clear expiration date. The mood among the AI community shifted sharply over the weekend, from frustration to a deeper unease that the era of wide-release frontier models may be over for good.

Mythos Returns, But Only for a Hundred Partners

Commerce Secretary Howard Lutnick sent a letter to Anthropic’s Chief Compute Officer Tom Brown—not CEO Dario Amodei—outlining a narrow path forward. Lutnick wrote that Anthropic had “worked with the U.S. government to address risks associated with Claude Mythos V and Claude Fable,” and that “appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos V model.” The Department of Commerce has not released the full list of approved organizations, but reports suggest roughly one hundred partners, including both private companies and U.S. government agencies, will regain access.