Anthropic lifts export restrictions: Claude Mythos 5 gains access to 100 US organizations, Fable 5 on the way
The U.S. Department of Commerce has officially lifted the ban on the export of Anthropic's Claude Mythos 5 model. This decision, made on Friday, opens access to the advanced AI system for more than 100 organizations — including major private companies and government entities listed in Appendix A of export controls.
Two-Week Standoff Concluded
The unblocking resulted from intense negotiations between the Anthropic team and the presidential administration. Commerce Secretary Howard Lutnick confirmed in a letter to Anthropic's CTO Tom Brown that current safety measures are deemed sufficient for transferring the model to verified partners — without the need for additional licenses.
Recall that both products — Mythos and Fable — were previously disabled after one of Anthropic's key investors, Amazon, expressed concerns about their safety. Company researchers warned that Fable 5 could be hacked for potentially dangerous purposes.
Mythos 5: From Glasswing to Broad Access
Until now, Mythos was used exclusively within the Glasswing program — a vulnerability discovery initiative involving about 150 organizations from more than 15 countries. Notably, the model has already proven its effectiveness: during government testing, Mythos identified errors in classified systems within hours.
Now, the restrictions that hindered its launch have been fully lifted. This means U.S. companies and agencies will gain access to a tool capable of analyzing critical infrastructure for vulnerabilities.
Fable 5: Waiting and a New Regulatory Landscape
The fate of Fable 5 remains less certain. Unlike Mythos, access to this model was previously open to all subscribers, and for a time it was considered the most powerful public AI tool. However, its release process is now stalled, and exact timelines have not been announced.
The situation is compounded by the formation of a new system of restrictions. On June 2, an executive order was signed creating a voluntary channel for federal review of advanced AI models. Developers can submit their solutions for cybersecurity checks 30 days before launch. Simultaneously, Washington is tightening AI chip exports to China, and now similar requirements are being extended to access the models themselves.
OpenAI followed the same course: on Friday, the company restricted access to the most powerful version of GPT-5.6, called Sol, to about 20 partners who received government approval, while the weaker versions Terra and Luna became publicly available.
The initial reason for the blocks was concern that the technologies could be obtained by competitors from China. Attention was focused on South Korean operator SK Telecom, which was added to the Glasswing list in early June and then had its access restricted. SK Telecom denied any ties to China.
Leaders of major cybersecurity companies, including former Facebook security chief Alex Stamos, as well as representatives from Nvidia, Adobe, and Zoom, urged authorities to abandon restrictive measures. Now that Mythos 5 has received the green light, the question is whether Fable 5 will follow.
My analysis: The lifting of restrictions on Mythos 5 is not just a technical decision, but a signal of the formation of a new AI regulatory model in the U.S. Washington is clearly seeking to maintain control over the most powerful models, granting access only to a limited circle of trusted entities. For the crypto industry, this means that access to advanced AI tools for analyzing smart contracts and DeFi protocols could become a privilege rather than a right. Fable 5 will likely receive similar approval, but under even stricter conditions.