On Friday, the U.S. Department of Commerce officially lifted the export ban on Anthropic's Claude Mythos 5 model. This decision opens access to one of the most powerful AI models for over 100 American organizations, ranging from the largest corporations to government agencies. The two-week standoff between the Trump administration and Anthropic ended in the developer's favor: Mythos 5 is returning, while Fable 5 remains off-limits for the general public for now.

Department of Commerce Approves Mythos 5 for Verified Partners

Commerce Secretary Howard Lutnick personally notified Anthropic's Chief Technology Officer Tom Brown about the lifting of restrictions. According to the letter, a license is no longer required to transfer the model to companies and agencies on the Appendix A list. "I have concluded that the existing safeguards allow certain verified partners to access the Claude Mythos 5 model," Lutnick stated.

Recall that both models — Mythos and Fable — were previously disabled after Anthropic's key investor, Amazon, expressed concerns about their safety. Company researchers warned that Fable 5 could be hacked for potentially dangerous purposes. Until now, Mythos was only used within the Glasswing project, a vulnerability discovery program involving approximately 150 organizations from over 15 countries. Notably, within just a few hours of testing, Mythos identified critical errors in classified systems.

Fable 5 Still Awaits: New AI Rules Are Only Taking Shape

According to sources close to the negotiations, the process for releasing Fable 5 has moved forward, but no exact timeline has been announced yet. Unlike Mythos, access to Fable 5 was previously open to all subscribers, and for a time it remained the most powerful public AI tool. The situation is gradually evolving into a new system of restrictions: on June 2, an executive order was signed, opening a voluntary channel for federal review of advanced AI models. Developers can submit their solutions for cybersecurity checks 30 days before launch.

Interestingly, OpenAI followed the same path on Friday. The company restricted access to the most powerful version of GPT-5.6, called Sol, to approximately 20 partners who received government approval. Weaker versions, Terra and Luna, became available to the general public. Initially, such blockages were driven by concerns that China might gain access to the technologies. Media reported that attention was linked to SK Telecom, a South Korean operator added to the Glasswing list in early June, after which its access was restricted. SK Telecom denied any connection to China.

Many leaders of major cybersecurity companies urged authorities to abandon restrictive measures. An open letter organized by former Facebook head of security Alex Stamos was signed by leaders from Nvidia, Adobe, Zoom, and other companies. Countries in Europe and other regions note that they have suddenly become heavily dependent on Washington's decisions regarding access to new developments. Whether Fable 5 will receive similar permissions will become clear in the coming days.

My comment: Lifting export restrictions on Mythos 5 is not just a victory for Anthropic, but a signal of a new paradigm in AI regulation. We are witnessing a shift from total bans to targeted access, which could set a precedent for all future models. Fable 5 will likely receive similar approval, but with even stricter conditions — the market is preparing for the segmentation of AI tools by trust level.