Austria Wants to Bring Anthropic to Europe

Austria is pushing the European Union to examine settling the US company Anthropic within the bloc. In a letter to EU Technology Commissioner Henna Virkkunen, released by the Austrian government, State Secretary for Digitalisation Alexander Pröll (ÖVP) proposes jointly exploring the “strategic establishment and participation of Anthropic within the European Union” – with legal certainty, market access, capital and a set of values suited to the company.

According to a spokesperson for Pröll speaking to news agency APA, the aim is to strengthen Austria as a location and to make Europe more digitally sovereign. The State Secretary left open how such an establishment might work in practice. He also acknowledged that there would be scepticism about whether the step is feasible at all.

The Trigger: US Export Controls on Fable 5 and Mythos 5

The proposal stems from a conflict between Anthropic and the US government. In early June, the company released its most powerful models to date, Fable 5 and Mythos 5, only to shut down access worldwide a few days later. The reason was an export-control directive from the US Commerce Department barring any use by foreign nationals – regardless of location and including Anthropic’s own non-US employees. Because filtering users by nationality in real time was effectively impossible, Anthropic disabled both models for all customers. Access to the other Claude models remained unaffected.

Legally, this is a licensing requirement rather than a permanent ban: the responsible Bureau of Industry and Security invoked the Export Control Reform Act and requires an individually validated export licence before foreign nationals may be granted access. According to reports, it was the first time the instrument had been applied to a commercially available AI model. Anthropic described the move as a misunderstanding and entered negotiations to restore access. In late June, US authorities granted limited clearance for wider use of Mythos 5 by vetted partners.

“A Profoundly European Stance”

In his letter, Pröll sets out why he believes this particular company would be a fit for Europe. Anthropic, he argues, treats the ethical use of AI not as marketing but as a core conviction, and prioritises safety over speed – an attitude he describes as profoundly European. In Europe, Pröll writes, the company would not be constrained but “unleashed”.

The State Secretary himself addressed the fundamental objection that such a venture is hardly realistic. “The real question is not whether it is easy,” he wrote. What matters, in his view, is whether Europeans are prepared to be the architects of their own technological future – or to remain mere administrators of decisions taken elsewhere.

Part of the Sovereignty Debate

The proposal lands at a moment when the EU is seeking to reduce its dependence on US and Chinese technology. In early June, the European Commission presented a technological sovereignty package, including a Cloud and AI Development Act intended to strengthen the domestic cloud, AI and semiconductor industries. The US restrictions on Fable 5 and Mythos 5 had immediately triggered calls for greater digital sovereignty across several member states.

Such considerations have gained urgency since US President Donald Trump took office. Initially it was above all his tariff policy that exposed the extent of European dependence; the access block on Anthropic’s top models then made clear the degree of control Washington holds in the AI field. At the same time, Europe still has no model that can match the most capable US systems.

Anthropic in Repeated Conflict With Washington

That a move to Europe is being discussed at all reflects the recurring tensions between Anthropic and the US government. Even before the release of its most capable model, a dispute between the Pentagon and the company over permissible AI uses had escalated: because Anthropic had drawn red lines, for instance on autonomous weapons systems, the US Defense Department classified the firm as a risk – a designation Anthropic is challenging in court. The case remains pending.

At the same time, there are indications that Anthropic continues to cooperate with US authorities in other areas. The relationship is therefore less a clean break than a shifting mix of conflict and collaboration.

How Realistic Is the Proposal?

A number of observers consider an establishment unlikely. On the decisive point – fresh capital – Europe cannot yet compete with the United States. Despite its friction with the government, Anthropic is currently preparing a US stock-market listing and has already filed a confidential application. A switch to Europe would also risk the company’s access to its by far most important market; a political attempt by Washington to block such a step is likewise seen as conceivable.

Assessments of the strategic benefit also diverge. Some experts argue that Europe is, for now, not significantly worse off without access to the two models, and that a restoration of access is likely. Others point out that Anthropic itself described its top models as especially dangerous in its own communications – which is what enabled the regulatory intervention in the first place. From this perspective, beyond questions of location, there would also be security-policy and legal hurdles to resolve.

Anthropic did not initially respond to a request for comment on the Austrian proposal. Whether and in what form the idea will be taken up in Brussels remains open.