Anthropic finally lifted a U.S. export ban that had shut its Claude Mythos 5 and Claude Fable 5 behind a wall. The restriction stemmed from a 2024 directive that followed Amazon CEO Andy Jassy’s warning about a jailbreak exposing model misuse. Weeks of talks with Commerce Secretary Howard Lutnick brought the first easing for customers.

Lutnick’s letter, sent Friday, said “appropriate safeguards are in place to permit certain trusted partners to access Claude Mythos 5.” The government cleared the model for over 100 U.S. institutions, ranging from Fortune 500 firms to federal agencies. The decision comes just after OpenAI rolled out GPT‑5.6 to a limited partner cohort for developers today again.

Yet the release leaves Claude Fable 5 still off‑limits. Lutnick’s note was silent on the second model, though insiders say a rollout could follow. The partial lift signals a shift in how the U.S. balances emerging AI power with national‑security concerns, setting a precedent for future model approvals in the tech world today again now.

For the 100 institutions slated to receive Claude Mythos 5, the model promises higher‑resolution reasoning and safer guardrails, potentially reshaping enterprise AI workflows. The government’s decision underscores that regulatory gatekeeping can coexist with commercial deployment, but it also hints that any future releases will be tightly vetted across the industry today again now.