The U.S. Department of Commerce has officially approved the removal of export restrictions for two flagship Anthropic models — Claude Fable 5 and Mythos 5. The decision, made on June 30, takes effect on July 1, meaning the immediate restoration of full global access to these artificial intelligence systems.

This marks the culmination of nearly three weeks of intensive negotiations between Anthropic and White House representatives. As a reminder, the restrictions were imposed on June 12, just three days after the models' release on June 9. Authorities cited national security risks related to potential attempts to bypass protective mechanisms. In response, Anthropic completely blocked access to both models for all foreign users across all its platforms, including Claude.ai, the API, and partner services like AWS Bedrock.

Why were the restrictions lifted?

The issue turned out to be less about the models themselves and more about the technical impossibility of promptly filtering users by nationality in real time. Anthropic took an unprecedented step by completely disabling Fable 5 and Mythos 5 for all clients, which effectively paralyzed their commercial use. The U.S. administration, realizing that such measures harm the competitiveness of American technology, made concessions.

"We have received notification that the U.S. Department of Commerce has lifted export restrictions on Claude Fable 5 and Mythos 5. Starting tomorrow, we will begin restoring access and will share details soon," Anthropic confirmed on June 30.

What does this mean for the AI market?

Claude Fable 5 is Anthropic's advanced model, built on the Mythos architecture but with additional fine-tuned safety features for operation across a wide range of scenarios: from programming to image analysis. Mythos 5, in turn, was originally designed for more sensitive tasks, including cybersecurity, and access was limited to select partners through the Project Glasswing program for critical government and corporate projects.

Pricing remains competitive: $10 per 1 million input tokens and $50 per 1 million output tokens. Anthropic's internal classifiers automatically redirect high-risk queries to safer algorithms, particularly in the areas of cybersecurity and biology, across all company products.

My expert conclusion:

The lifting of restrictions is not just a victory for Anthropic, but a signal of the beginning of a more mature and pragmatic approach to AI regulation in the U.S. The administration has realized that harsh "early-stage" blockades kill innovation rather than protect against it. Now the market sees that dialogue between developers and regulators can be productive, and this is a positive precedent for the entire industry.