The White House ordered Anthropic to block all foreign access to its newest AI models Fable 5 and Mythos 5 on national security grounds, forcing the company to shut them down entirely—a move that signals U.S. government control over frontier AI technology and may accelerate other countries' push to develop their own.

What happened

On June 12th, following reports that Amazon security research found ways to use Fable 5 to obtain information usable in cyberattacks, the White House issued an order requiring Anthropic to block access to Fable 5 and Mythos 5 for all foreign nationals, including Anthropic's own foreign-born employees. Anthropic complied by completely cutting off access to both models for all customers. The company said it disagreed with the decision, noting that the potential jailbreak was narrow and that the models had already been subject to safeguards limiting use in high-risk areas.

Why it matters

The sudden shutdown without detailed explanation demonstrates that the U.S. government can unilaterally restrict access to cutting-edge American AI technology on national security grounds—a power that may concern countries and companies relying on U.S. technology leadership. The incident has already provided ammunition to politicians, governments, and companies arguing they need to lead in AI technology themselves rather than depend on American firms.

What to watch

Reports indicate the White House's decision was partly driven by concerns that a group linked to China had accessed Mythos 5, raising the possibility of reverse-engineering through distillation (training a simpler AI model on a more advanced one to copy its behavior). The White House has not publicly confirmed this or provided specific details of its national security concerns to Anthropic.