Anthropic Cuts Off Foreign Users: The Start of the AI Sovereignty Race

On June 12, 2026, acting on a directive from the US Department of Commerce, Anthropic shut off access to all of its models for foreign users, with no warning and no transition period.

How the shutdown unfolded

The directive arrived at 5:21 p.m. Eastern time on June 12, 2026. A few hours later, millions of users around the world opened their usual interface and saw an error message. Anthropic gave no grace period: no advance warning, no time to export data, and no chance to switch to an alternative.

The cutoff did not hit only corporate customers and independent developers. Even Anthropic's own foreign employees lost access, including people who had worked directly on the models that were shut off.

• June 12, 2026: date of the shutdown
• 5:21 p.m. Eastern time: when the US Department of Commerce directive arrived
• Fable 5 and Mythos 5: the Anthropic models that stopped working abroad
• All foreign users, including Anthropic employees, were blocked
• Transition period: none at all

Why this became the main argument for sovereign AI

The incident exposes a systemic vulnerability the industry has ignored for years: access to frontier AI models depends entirely on the decisions of a single national government. When the US Department of Commerce signs a directive, thousands of companies, researchers and government bodies worldwide lose, that same evening, a tool around which they built products and business processes.

Before Fable 5, similar restrictions targeted hardware: a number of countries have long been unable to legally buy Nvidia H100 and H200 chips. But a chip ban affects future investment, which is painful but not instant. Blocking a language model hit differently: it struck finished products, live workflows and employees, in the middle of the working day, with no chance to do anything about it.

Sovereign AI is not national pride but infrastructure insurance: models trained on local data, running on local servers and answerable only to a local regulator. After June 12, 2026, that argument is no longer theoretical.

"It is hard to imagine a more persuasive advertisement for sovereign

AI," is how the authors of the original piece describe what happened.

The market is reacting predictably. Europe is speeding up funding for national AI programs: Mistral in France, Aleph Alpha in Germany and similar projects have gained an extra political argument for state investment. China keeps expanding capacity in its own ecosystem, which has been effectively separate from the Western one for several years. Russia, the UAE, India, Japan and South Korea are investing in models that an outside regulator cannot unilaterally recall. The global AI model market is becoming multipolar, not for ideological reasons but out of pure pragmatism.

What this means

Anthropic's June 12, 2026 shutdown is not an isolated incident but a symptom of a structural break in the global AI industry. The market is splitting into clusters: American flagship models under strict export controls, Chinese models under their own restrictions, and a fast-growing field of regional and sovereign alternatives. For companies and states, the choice has become plain: build their own AI infrastructure, or accept a dependency that a foreign government can cut off without warning, at any moment, on any day.

What happened to Claude on June 12?

Anthropic cut off access to Fable 5 and Mythos 5 for foreigners after a directive from the US Department of Commerce. The shutdown came a few hours after the directive arrived at 5:21 p.m. Eastern time on June 12, with no transition period.

Can users in Russia use Claude?

According to the article, on June 12, 2026, Anthropic cut off access for all foreign users, including those in Russia. Users got an error message when they tried to connect.

Which Anthropic models are disabled for foreigners?

Under the US directive of June 12, all Anthropic models are disabled, including the new Fable 5 and Mythos 5. Millions of users received an error message within a few hours of the order.

Why did

Anthropic comply with the US directive without a transition period? The directive came from the US Department of Commerce at 5:21 p.m. Eastern time. The company had to comply immediately, with no time to export data or warn users.