US export control impact on AI: Europe locked out of Anthropic models

When the US government sent Anthropic a letter at around 5:21 p.m. ET on a Friday, citing national security concerns, few expected the fallout to reach Brussels by Sunday. Yet the US export control impact on AI was swift and far-reaching — forcing Anthropic to pull its most advanced models entirely and prompting the European Commission to open an urgent assessment of what that means for European users and European digital autonomy.

Key takeaways

• The Trump administration ordered Anthropic to suspend foreign-national access to its most advanced AI models, Fable 5 and Mythos 5, citing national security concerns.
• Anthropic said it would disable those models for all users — including its own foreign-national employees — to ensure compliance with the export control directive.
• The European Commission confirmed on June 14, 2026 it is assessing the practical consequences of the move for European users of those services.
• Commission spokesperson Thomas Regnier said contingency measures should not be discriminatory against partners and called for Europe to strengthen its technological sovereignty.
• The directive arrived with no advance warning and did not provide specific details of the underlying national security concerns, according to Anthropic.

Anthropic Cuts Off Its Most Capable Models Following Government Order

The move came without warning. Anthropic received a government letter Friday evening ordering it to suspend access to Fable 5 and Mythos 5 — two of the most capable AI models on the market — for all foreign nationals, whether inside or outside the United States. That definition, the company said, explicitly included its own foreign-national employees: the very engineers who helped build those systems.

Faced with a directive that applied so broadly, Anthropic concluded the only way to comply was to disable the models for everyone. The letter cited national security concerns but, according to the company, did not provide specific details about the underlying threat.

The decision sent immediate shockwaves through the AI community. Experts, researchers, and executives reacted with a mix of alarm, skepticism, and outright disbelief. Dean W. Ball, a senior fellow at the Foundation for American Innovation, called it “simply cartoonish,” pointing to the apparent contradiction of an administration willing to export advanced AI chips while simultaneously banning its closest allies from using American AI models. Gary Marcus, a prominent AI researcher, warned the decree was “wildly overdramatic and counterproductive for the US AI industry,” adding that it risked pushing Chinese engineers working at US labs to return home.

The export control’s reach turns on a legal concept known as the “deemed export” rule — a provision under which showing controlled technology to a foreign national inside US borders counts legally as exporting it abroad. As Peter Girnus, a senior threat researcher at Zero Day Initiative, put it bluntly: “The munition is in the building and the people who made it are not allowed to look at it.”

European Commission Steps In to Assess the Damage

Brussels moved quickly. On June 14, 2026, the European Commission confirmed it was actively examining the practical consequences of the US directive for European users of Anthropic’s services. The statement came from Commission spokesperson Thomas Regnier, and it was pointed.

“We are looking closely at the practical consequences of this for European users of these services,” Regnier said.

The Commission’s concern is not abstract. European businesses, researchers, and public institutions that rely on Anthropic’s flagship models now find themselves cut off by a unilateral American national security decision that was made with no consultation and no advance notice. The disruption is real and immediate.

AI Benefits and Cybersecurity Risks Acknowledged

Regnier did not frame the Commission’s response as purely defensive. He acknowledged that a new generation of highly capable AI models carries genuine dual-use complexity. “These models offer significant benefits, including for cyber-defence, but they also raise serious cybersecurity concerns that need to be addressed,” he said.

That framing matters. It signals that the European Commission is not dismissing the underlying logic of the US directive entirely — it understands that frontier AI systems pose real security considerations. What Brussels objects to is the method, not the principle.

A Clear Line on Non-Discrimination

On that point, the Commission was unambiguous. “We believe that contingency measures taken in this light should not be discriminatory against partners,” Regnier said. The word “partners” carries weight here. The European Union is among the United States’ closest allies. Measures framed around national security that sweep up partner nations indiscriminately sit uncomfortably within any conventional understanding of alliance relationships.

Chris McGuire, a senior fellow for China and emerging technologies at the Council on Foreign Relations, made a similar point from within the US debate: “Targeted export controls on model access are prudent. But across the board controls on all countries on a single model, without any warning, is highly questionable.”

The Bigger Picture: Europe’s Technological Sovereignty

Perhaps the most significant element of Regnier’s statement was its forward-looking dimension. “This development is a further illustration of why Europe needs to strengthen its technological sovereignty,” he said.

That sentence carries the full weight of a longer European anxiety. For years, European policymakers have warned about strategic dependence on American technology platforms. The AI era has intensified those concerns considerably. When a single US government letter can overnight disable access to frontier AI models for European users, researchers, and businesses — with no prior notice, no consultation, and no appeal mechanism — the argument for European AI independence becomes harder to dismiss.

The incident also raises a broader question about the direction of US AI regulation. Aaron Levie, CEO of Box, described it as “a big turning point for AI regulation,” noting that the government is beginning to deem some models too powerful for certain uses — a precedent that could shape a wide range of controls going forward. Alan Rozenshtein, research director at The Lawfare Institute, went further, suggesting the case “may become the first big First Amendment AI case.”

What the European Commission’s assessment will ultimately recommend remains unclear. No timeline has been given, and the specific measures under consideration have not been disclosed. But the political signal is unmistakable: Europe is treating this not as a one-off incident, but as a structural problem — one that demands a structural response. The race to build competitive European AI capacity just got a sharper geopolitical edge.

FAQ

Why is Anthropic disabling access to its advanced AI models?

Anthropic is disabling its most advanced AI models, Fable 5 and Mythos 5, following a US government export control directive that cites national security concerns. The order requires suspending access for all foreign nationals — including Anthropic’s own foreign-national employees — and the company determined that the only way to ensure compliance was to disable the models for all users.

What is the European Commission’s stance on the US export control measure?

The European Commission confirmed on June 14, 2026 that it is assessing the practical implications of the US directive for European users. Spokesperson Thomas Regnier stated that any contingency measures taken should not be discriminatory against partners, and that the situation underlines the need for Europe to strengthen its technological sovereignty.

What are the European Commission’s concerns regarding AI technology in this context?

The Commission recognizes that highly capable AI models offer significant benefits — including for cyber-defence — but also raise serious cybersecurity risks that need to be addressed. The broader concern is Europe’s strategic dependence on US-controlled AI infrastructure, and the Commission is treating the incident as a catalyst for accelerating efforts toward greater European AI technological sovereignty.

Article produced with the assistance of artificial intelligence and reviewed by the editorial team.