U.S. Government Partially Restores Access to Anthropic's Mythos 5 AI Model

The U.S. Department of Commerce issued a letter on June 26, signed by Secretary Howard Lutnick, that partially reverses the export ban imposed on June 12 on Anthropic’s advanced AI models. The new policy permits more than 100 trusted American organizations—including major tech firms and federal agencies—to access the Claude Mythos 5 model without a licensing requirement, while still restricting redistribution to non‑U.S. entities and employees.

The public‑facing version, known as Fable 5, remains blocked for all users worldwide. The original shutdown was driven by national‑security concerns over potential model jailbreaks and alleged Chinese interest. Negotiations continue, and Anthropic expects the block on Fable 5 could be lifted later this week, though no firm date has been set. The same day, OpenAI released a limited‑access GPT‑5, highlighting a broader trend of U.S. regulators imposing selective rollout rules for frontier AI systems.