US blocks foreign access to Anthropic's advanced AI models for national security
The Trump administration issued an export control directive on Friday ordering Anthropic to suspend access to its newest and most powerful AI models, Fable 5 and Mythos 5, for all foreign nationals, whether inside or outside the United States. The directive came just three days after Anthropic publicly released these models, which the company had marketed as state-of-the-art. In response, Anthropic disabled both models entirely for all customers to ensure compliance, affecting paying enterprise clients and even the company's own foreign-born employees. The government cited national security concerns, noting it had learned of a method to bypass Fable 5's safety guardrails. Anthropic disputes the severity, arguing the jailbreak method is narrow and similar vulnerabilities exist in other publicly available models not subject to such restrictions. The company called the action a misunderstanding and said other Anthropic models remain unaffected. This marks an unprecedented escalation in treating advanced AI systems as controlled national security assets comparable to semiconductors.
Key facts
- US Commerce Department issued export control directive Friday evening citing national security authorities
- Directive prohibits access to Fable 5 and Mythos 5 by any foreign national inside or outside US, including foreign Anthropic employees
- Models were released publicly only June 9, just three days before the directive
- Anthropic disabled models entirely for all customers to comply, affecting enterprise clients globally
- Government cited awareness of a jailbreak technique bypassing Fable 5 safety guardrails
- Anthropic claims the jailbreak targets narrow capabilities also found in OpenAI's GPT-5.5
- This is first export control directive on AI models themselves, not hardware or chips
- The action escalates tensions between Anthropic and Trump administration over military AI use restrictions
How different outlets framed it
Left-leaning coverage: Progressive critics view this as bullying of Anthropic for refusing military demands for surveillance and autonomous weapons. They argue the company is being singled out for defending responsible AI development and human oversight, with the directive representing an authoritarian overreach that contradicts American values.
Right-leaning coverage: Conservative outlets frame the restriction as prudent national security policy recognizing that advanced AI is a strategic weapon in geopolitical competition. They emphasize that Anthropic bears responsibility for describing its product as dangerously powerful, while supporting government protection of critical infrastructure and hardening against foreign threats.
Limits of this briefing
The Trump administration did not provide detailed national security justifications in writing. The specific jailbreak method and extent of actual security risk remain unclear from public statements. Future policy direction depends on ongoing negotiations between Anthropic and government officials.
Sources
- Bloomberg: Anthropic Says US Limits Foreign Access to Fable 5, Mythos 5 Centre
- Newsweek: Trump's ban on foreign access to new Anthropic models sparks tech backlash Right
- Al Jazeera: US orders Anthropic to disable AI models for all foreign nationals Centre
- Conservative Playlist: Trump Administration Blocks Anthropic's New Advanced AI Models as Too Dangerous for Foreign Hands Right
- Center for American Progress: The Trump Administration Is Trying To Make an Example of the AI Giant Anthropic Left