U.S. Blocks Foreign Access to Anthropic's Mythos, Fable 5 Models

The U.S. government has ordered Anthropic to immediately suspend access to its most advanced AI systems for all foreign nationals, marking an unprecedented federal intervention in how an AI company can deploy its technology.

Anthropic disclosed late Friday that the Commerce Department had directed the company to block access to its Fable 5 and Mythos 5 models, citing national security concerns. The order affects not only international customers but also foreign nationals employed by the San Francisco-based company.

Scope and Implementation

In a social media post, Anthropic stated it "must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance." The directive came from the Commerce Department, according to a person familiar with the matter who requested anonymity. The duration of these restrictions remains unclear.

The breadth of the order is striking. While export controls on technologies like advanced semiconductors to countries such as China are routine, this directive could theoretically prevent Anthropic employees who are citizens of allied nations—including Canada and the United Kingdom—from working on the affected AI systems.

Why it matters

This action represents the most direct federal control over AI model deployment to date, establishing a precedent for how the government might regulate access to frontier AI systems it deems sensitive. For enterprise leaders evaluating AI partnerships, the incident underscores a new category of regulatory risk: that advanced models could be suddenly restricted based on national security determinations, regardless of commercial agreements or operational dependencies.

Escalating Federal Tensions

The access restrictions mark the second time this year that federal authorities have targeted Anthropic. In March, following contentious negotiations over potential military and intelligence applications of the company's AI technology, the Pentagon designated Anthropic as an unacceptable supply chain risk. That classification could limit the company's ability to work with federal agencies.

The relationship between Anthropic and the federal government has grown increasingly strained, contrasting with the company's founding mission to develop AI safely. The latest order suggests authorities view certain capabilities in the Mythos and Fable 5 systems as sufficiently powerful or dual-use to warrant unprecedented access controls.

Unanswered Questions

Key details remain unclear, including the specific capabilities that triggered the order, whether the restrictions apply to all foreign nationals uniformly or target specific countries, and what criteria Anthropic must meet to restore access. The company has not elaborated beyond its initial disclosure.

For the AI industry, the order raises fundamental questions about how national security concerns will shape the development and distribution of increasingly capable systems, and whether similar restrictions might extend to other frontier AI companies.

The details were first reported by The New York Times.