US Orders Anthropic to Block Foreign Nationals from Claude AI

Government directive shuts down access to latest Claude models

Anthropic announced Friday night that the US government ordered the company to suspend all foreign nationals—including its own employees—from using its most advanced AI products. To comply, Anthropic disabled customer access to Fable 5 and Mythos 5, the newest versions of its Claude AI system, according to Mother Jones.

The company said the government cited national security concerns but provided no additional explanation for the directive. Anthropic had released Fable 5 as the first publicly available version of its Mythos model family, which features enhanced software engineering and visual understanding capabilities compared to earlier iterations.

Why it matters

This order represents a significant escalation in government control over AI deployment and marks the first known instance of the US forcing a private AI company to restrict access based on user nationality. The move could accelerate the global trend toward "sovereign AI," where countries build independent AI infrastructure to avoid foreign restrictions—potentially fragmenting the technology landscape and limiting cross-border collaboration on AI safety research.

Background of Trump-Anthropic conflict

The directive continues a months-long confrontation between the Trump administration and Anthropic. In February, the administration banned all federal agencies from using Anthropic products and Trump publicly called the company "radical left" and "woke."

The dispute centers on Anthropic's refusal to allow military applications that CEO Dario Amodei said would violate the company's safeguard policies. Specifically, Amodei stated the government demanded capabilities for mass surveillance of Americans and fully autonomous weapons systems.

Despite this stance, reporting from March revealed that the US military had previously used Claude for intelligence assessments, target identification, and battle scenario simulations during preparations for strikes on Iran.

Safety measures and market position

Anthropic has built its brand around ethical AI development, releasing a preview version of its latest model to select industry partners in April to test for potential misuse in creating hacking tools. The company implemented guardrails in Fable 5 that block responses to questions about cybersecurity, biology, and chemistry.

The company filed for an initial public offering earlier this month. The recent success of SpaceX's stock market debut—which made Elon Musk a trillionaire—could signal strong investor appetite for major AI companies, including Anthropic and competitor OpenAI.

Meanwhile, countries including China and the United Arab Emirates are accelerating development of their own AI infrastructure to reduce dependence on US-based systems and their associated data privacy and safeguard requirements.

These details were first reported by Mother Jones.