Anthropic says the U.S. Department of Commerce has lifted export controls that halted deployment of its Claude Fable 5 and Mythos 5 models, ending more than two weeks of uncertainty.
Beginning Wednesday, the company will restore Fable 5 through Claude.ai, Claude Code, Amazon Web Services, Google Cloud, Microsoft Foundry, and selected enterprise offerings. Mythos 5 is also returning to approved organizations through Anthropic’s Glasswing cybersecurity program.
The dispute began days after Anthropic introduced Fable 5, one of its most capable public models. Federal officials directed the company to suspend access while reviewing national security concerns tied to the release.
With the restrictions lifted, Anthropic can resume its rollout plans, closing one of the first major government reviews of a frontier AI system under the federal voluntary review framework.
Why It Matters: The outcome provides an early example of how advanced AI systems may move from development to public release under federal review. Product launches and government oversight are becoming more closely connected, adding a new consideration for companies building and deploying frontier AI.
- Federal Review Enters the Product Launch Process: Anthropic suspended Fable 5 after receiving a Commerce Department directive that blocked access for foreign nationals. The company resumed deployment after the government approved additional safeguards, offering one of the first examples of a frontier AI model moving through formal federal review before wider release.
- Security Reviews Become Part of Model Development: Government concerns reportedly centered on whether Fable 5 could be “jailbroken” to bypass its safeguards. Anthropic argued the identified vulnerabilities were limited and could also be reproduced with other frontier models. Security assessments now appear to play a larger role in determining when advanced AI models are ready for public release.
- The Pause Raises Questions About U.S. Competitiveness: While Fable 5 remained unavailable, technology executives and investors warned that the delay gave Chinese open-source developers more time to close the gap with leading U.S. models. Commerce Secretary Howard Lutnick said the approval supports national security while helping maintain American leadership in AI.
- Washington and AI Companies Work More Directly Together: Anthropic spent weeks negotiating with Commerce Department officials, with co-founder Tom Brown reportedly leading the discussions. The company joins OpenAI in participating in the federal government’s voluntary review process for advanced AI models, suggesting that coordination with Washington could become a more common part of major model releases.
- Policy Questions Remain Open: The restoration of Fable 5 and Mythos 5 resolves this dispute, but debates over export controls, AI safety standards, military applications, and long-term oversight continue. Anthropic also remains engaged in separate legal and policy disputes with the federal government, leaving many questions about frontier AI governance unanswered.
Trusted insights for technology leaders
Our readers are CIOs, CTOs, and senior IT executives who rely on The National CIO Review for smart, curated takes on the trends shaping the enterprise, from GenAI to cybersecurity and beyond.
Subscribe to our 4x a week newsletter to keep up with the insights that matter.
