The United States government has blocked Anthropic from exporting two of its most powerful artificial intelligence models — Mythos 5 and Fable 5 — marking what observers are calling the first-ever government ban on a specific AI model's export. The move, citing national security concerns, has set off an urgent scramble inside the San Francisco-based AI company.
According to the Times of India, the US government sent a formal letter directly to Anthropic CEO Dario Amodei warning that he could face civil penalties in connection with the ban. The letter named both Mythos 5 and Fable 5 explicitly.
In response, Anthropic dispatched a researcher to engage directly with government officials over AI safety concerns, according to Crypto Briefing. The company is actively seeking to regain access to the blocked models.
Those efforts have not yet paid off. According to MSN, citing reporting on Anthropic's negotiations with the Trump administration, there has been "no breakthrough" in talks so far. The company is described as working to have the restrictions lifted, but the two sides remain at an impasse.
The ban is significant because it represents a new frontier in how governments may seek to regulate frontier AI: not just controlling chips or data centers, but the models themselves. If the approach spreads, it could reshape how AI companies operate globally and who gets access to the most capable systems.