U.S. Bars Foreigners From Using Anthropic's Most Advanced AI Models

The United States government has ordered Anthropic to block all foreign nationals — including those physically inside the U.S. and even employees working at Anthropic itself — from accessing the company's most advanced artificial intelligence models, Fable 5 and Mythos 5, citing national security concerns.

In a blog post published Friday, Anthropic said it received a directive from U.S. government agencies at 5:21 PM ET ordering the company to immediately cut off access to its cutting-edge AI systems for anyone who is not a U.S. citizen or permanent resident. The company confirmed it has complied with the order, disabling access to both models globally for foreign users.

Why the Government Stepped In

The artificial intelligence powering Anthropic's Mythos model is particularly adept at detecting software vulnerabilities — including some that have remained undiscovered for decades. This capability has been used by U.S. authorities and selected corporate partners to identify and patch security gaps in critical infrastructure. However, policymakers have grown increasingly concerned that such powerful AI could become a dangerous cyberweapon if it fell into the wrong hands.

Fable 5, which Anthropic released just this week as a commercial product based on Mythos technology, includes cybersecurity and biotechnology capabilities that the government determined should be restricted. Mythos 5, the non-public full version, will continue to be used exclusively by government agencies and approved partners to harden their systems.

"We received a letter at 5:21 PM on Friday directing us to prevent all foreign nationals from accessing our newly released models. The letter did not explain the government's specific security concern in detail." — Anthropic blog post

Anthropic said it disagreed with the order, arguing that the safety measures in Fable 5 had been extensively tested and that the software, already used by hundreds of millions of users, should not have been blocked for this reason. The company noted it had received only partial information from the government about the specific capabilities that triggered the restrictions.

Broader Context

The order represents one of the most aggressive actions yet by the U.S. government to restrict access to frontier AI capabilities. Earlier this month, Anthropic itself proposed that the world's top AI companies coordinate to pause development of advanced systems, warning that the technology is improving so rapidly that there is a risk humans could lose control.

The ban also affects foreign nationals currently residing in the United States — including those working at Anthropic — as well as researchers and developers overseas who rely on these models for security research and software development. The company has now taken both Mythos 5 and Fable 5 offline for all non-U.S. persons, complying fully with the export control directive despite its disagreement with the rationale.

Sources: Al Jazeera, Reuters, The New York Times, Fortune