Anthropic Shuts Down Its Most Advanced AI Models After U.S. National Security Order

Anthropic said Friday that it is suspending access to its two most advanced models, Claude Fable 5 and Claude Mythos 5, after receiving a U.S. government directive tied to national security. The move came only days after Fable 5 was released to the public.

According to the company, the U.S. Commerce Department ordered it to restrict use of the models by any foreign citizen, whether inside or outside the United States, including Anthropic employees who are foreign nationals. Anthropic said it was not given full details of the security concerns behind the decision.

The company said it understood the concern to involve possible ways to bypass some of the model’s safeguards, which could make it useful for finding software vulnerabilities and carrying out advanced cyber operations. Rather than block only foreign users, Anthropic chose to pause both models for all users to ensure full compliance.

Anthropic criticized the government action, saying authorities should be able to stop unsafe deployments only through a transparent, fair, clear process based on technical facts, and that this order did not meet those standards. The company is also on the Pentagon’s blacklist, where its systems are considered too dangerous for government use, a decision it has sued the Trump administration over. The case may signal a broader shift in U.S. policy, with export restrictions now aimed directly at AI models rather than mainly at chips and hardware.