Anthropic breaks through the blockade: Mythos 5 gets the "green light" for 100+ organizations in the US, Fable 5 on the horizon
The U.S. administration has officially lifted the two-week ban on the export of Anthropic's flagship model, Claude Mythos 5. This decision opens access to the cutting-edge AI tool for over 100 American companies and government agencies included in the special "Appendix A" list. A separate license is no longer required to transfer the model—only verified partner status is needed.
Diplomatic Breakthrough and Security Lessons
The unblocking of Mythos 5 resulted from intense negotiations between Anthropic's top management and administration representatives in Washington. Commerce Secretary Howard Lutnick personally confirmed that current precautionary measures are deemed sufficient for the model to operate in a trusted environment.
Recall that both models—Mythos and Fable—were previously disabled after Amazon, a key investor in Anthropic, expressed serious concerns about their safety. Company researchers warned that Fable 5 is potentially vulnerable to hacking and could be used for dangerous purposes. Until then, Mythos itself was used exclusively within the Glasswing project—a vulnerability search program involving about 150 organizations from 15 countries. During government testing, Mythos identified critical errors in classified systems within hours.
Fable 5 Awaits Its Turn, and the Rules of the Game Are Changing
Unlike Mythos, the fate of Fable 5—previously the most powerful public AI tool—remains unclear. Sources close to the negotiations report that the release process has moved forward, but specific timelines are not disclosed. Previously, access to Fable 5 was open to all subscribers, making it unique in the market.
The situation unfolds against the backdrop of a new regulatory framework. 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 exports of AI chips to China, and the same is now happening with access to the models themselves.
Notably, OpenAI followed the same path on Friday, restricting access to the most powerful version of GPT-5.6, named Sol, to about 20 government partners, leaving the weaker versions Terra and Luna for the general audience.
Initially, the blockages were driven by fears that competitors from China could gain access to the technology. Attention 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 denies any connection to China.
Interestingly, many 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, European countries and other regions find themselves heavily dependent on Washington's decisions regarding access to new developments.
My analysis: The decision on Mythos 5 sends a clear signal to the market: the era of "open AI" is ending. We are entering a phase where access to advanced models will be strictly regulated by the state, and the software itself will become as strategic a resource as chips and rare earth metals. Investors and developers should prepare for the fragmentation of the AI market along national lines.