On June 30, the U.S. Department of Commerce officially lifted export restrictions on Anthropic's flagship artificial intelligence models — Claude Fable 5 and Mythos 5. This decision, reached after nearly three weeks of intensive negotiations between the company and the White House, takes effect immediately. Starting July 1, global access to both models will be fully restored.
As a reminder, the restrictions were imposed on June 12, just three days after the models' release on June 9. Authorities cited national security concerns related to potential attempts to bypass protective mechanisms. In response, Anthropic was forced to block access for foreign users across all its products, including Claude.ai, the API, AWS Bedrock, and partner services. The situation was complicated by the fact that real-time user filtering by nationality proved technically unfeasible.
What has changed?
The lifting of export controls is not merely a bureaucratic formality. It signals that dialogue between developers of advanced AI systems and regulators can be productive. Anthropic held a series of closed-door meetings with the Department of Commerce and administration officials, demonstrating the reliability of its security architectures. The result was swift: Commerce Secretary Howard Lutnick personally confirmed that after a thorough analysis, Fable 5 was deemed compliant with national security requirements.
Key details:
- Claude Fable 5 — Anthropic's most widely used and accessible model, built on the Mythos architecture with finely tuned protective mechanisms. It is designed for complex tasks, large-scale data processing, programming, and image analysis.
- Mythos 5 — a more sensitive model for critical applications, including cybersecurity and government projects. Initially, access was limited to trusted partners through the Project Glasswing program.
Pricing remains competitive: $10 per 1 million input tokens and $50 per 1 million output tokens. Anthropic's internal classifiers automatically redirect high-risk requests to safer algorithms, particularly in the fields of cybersecurity and biology.
Cryptalist Analysis
This precedent is an important step toward shaping mature AI regulation in the U.S. The swift lifting of restrictions not only reflects Anthropic's strong negotiating position but also indicates that authorities understand a complete ban on advanced models harms not only businesses but also the country's global technological leadership. However, investors and developers should closely monitor regulators' next moves: such "swings" could recur with the release of even more powerful models. The AI market is becoming an arena not only for technological but also for political competition.