Anthropic shutdown under Trump sparks non-American AI push

Anthropic took its most powerful models, Fable 5 and Mythos 5, offline at Washington's request over a weekend in June 2026 after the White House demanded the company block access for all foreign nationals, including its own employees. The shutdown followed government concern about who could reach frontier systems and came despite the models already being subject to safeguards limiting their use in "high-risk areas." Anthropic may bring Fable and Mythos back online, but the incident immediately sharpened debates about reliance on American AI.

What happened

The White House moved quickly and with little public explanation, ordering Anthropic to stop foreign access to Fable 5 and Mythos 5. The American company said it had little choice but to comply. Part of the administration's decision reportedly stems from a belief that a group linked to China had accessed Mythos, and Anthropic has accused Chinese rivals of using its models to train their own at scale. The shutdown removed access for foreign users, including Anthropic employees based abroad, and left many international customers cut off from the new models.

Global reactions and prior context

Leaders across democracies used the episode to press for greater domestic AI capability. In the UK, AI and online safety minister Kanishka Narayan used the shutdown to argue Britain must develop its own AI capacity, framing the issue as a matter of national security. "We treat every other threat to our sovereignty with deadly seriousness, but we haven’t learned to treat this one in the same way," he said.

In France the push was more explicit and political. Former Prime Minister Gabriel Attal called the shutdown the start of "the AI war" and likened the sudden withdrawal of access to a strategic chokepoint, saying it shows France's vulnerability if it relies on others for critical technologies. Members of the European Parliament and French media voiced similar alarm, and Attal said the issue will be central in France's upcoming presidential debates.

Canada's prime minister drew a similar lesson without naming wrongdoing. He argued the episode highlights the risk of relying on a single partner for crucial resources like AI and called for diversification.

Europe's unease is not new. The European Union has long pushed to reduce dependence on external providers across chips, cloud computing, and AI. The Anthropic shutdown made the problem feel immediate, adding to concerns about America's reliability as an ally under the current US administration.

Not all responses aim to match the scale of US or Chinese frontier labs. France's Mistral and Canada's Cohere are cited as examples of serious efforts outside the US and China, while Singapore and the UAE have targeted narrower priorities such as infrastructure or better local-language models. Open-source models are also mentioned as a potential path to capabilities that would be hard for any single party to control.

Why it matters

The shutdown exposed a simple strategic fact: access to frontier American models can be interrupted by US policy. That reality reframes AI as a component of national security and supply-chain risk, not just a commercial technology. Governments that depend on foreign models now face a choice between accepting potential future restrictions or investing in domestic and regional AI capacity. The episode strengthens political momentum for sovereign AI programs in Europe, Canada and elsewhere.

What to watch

Watch whether Anthropic restores global access to Fable 5 and Mythos 5 and what conditions the company or the White House attach to any return. Also track concrete policy moves: new funding or procurement decisions by European capitals, announcements from Mistral or Cohere about expanded capability, and any formal multilateral efforts to secure cross-border access guarantees for AI services.

Restoring customer trust in American AI will require more than bringing models back online. Many governments and companies saw how fragile access can be, and they are acting to reduce that fragility.