Summary

The US government under Donald Trump has blocked the use of Anthropic's newest AI models for non-Americans. Swiss EPFL Professor Marcel Salathé assesses this as a geopolitical weapon and warns of strategic dependencies. He calls for European millennium-scale investments in AI infrastructure, comparable to the NEAT major project. The move significantly impacts business and research in Europe, as advanced AI models are becoming increasingly critical.

People

  • Marcel Salathé (Professor, Swiss Federal Institute of Technology Lausanne)

Topics

  • Artificial Intelligence
  • Geopolitics & Technology
  • US Trade Policy
  • European Sovereignty

Clarus Lead

This move signals a technology Cold War: the US is using AI models for the first time explicitly as a geopolitical pressure tool against allies. For European decision-makers, this is a turning point – not just a regulatory issue, but an infrastructural dependency question. Salathé warns specifically: without massive European countermeasures, the continent will be permanently marginalized in critical technologies.

Detailed Summary

The Trump administration officially justifies the blockade with security concerns. Salathé decodes this as the deployment of AI infrastructure as a negotiation and pressure weapon in geopolitical conflicts. He compares the necessary European response to the NEAT major project – an infrastructure investment cycle that took decades but secured national independence.

The blockade has immediate effects: European researchers and companies lose access to the technologically leading models. This immediately creates an innovation gap. Salathé argues that Europe must not only replicate individual systems but build an independent AI ecosystem infrastructure – with European data standards, computing capacity, and models. This requires coordinated government investments in the billions over at least a decade.

Key Statements

  • The USA is using AI models for the first time as a geopolitical pressure tool – not just as technology
  • Europe must immediately launch massive infrastructure counter-investments to avoid dependency
  • The benchmark should be historical major projects: decade-long commitment, national sovereignty as the goal

Critical Questions

  1. Evidence (Security Argument): What specific security risks does the Trump administration cite for this blockade – and is there independent validation or is this pure rhetoric?

  2. Conflicts of Interest: Do US competitors of Anthropic (OpenAI, Meta) directly benefit from this market closure, and is there a hidden competitive interest at play?

  3. Causality: Is the comparison to NEAT (national infrastructure) actually analogous – or does Salathé underestimate the speed of AI development and the difficulty of European coordination on investments?

  4. Feasibility: What European mechanisms currently exist for such billion-dollar coordination – or would new governance structures need to be created?

  5. Alternatives: Can European companies continue to access Anthropic models through licenses, joint ventures, or third-country workarounds?

  6. Asymmetry: Does the blockade also affect Canadian or British users, and does this signal a broader US technology isolationism?

Bibliography

Primary Source: USA Block Anthropic AI Models – EPFL Professor Calls for Billion-Dollar Investments – Tages-Anzeiger (20.06.2026) https://www.tagesanzeiger.ch/anthropic-sperre-epfl-professor-fordert-eine-milliarde-fuer-schweizer-ki-257936463662

Authors: Joachim Laukenmann, Jan Bolliger

Verification Status: ✓ 20.06.2026

This text was created with the support of an AI model. Editorial Responsibility: clarus.news | Fact-Check: 20.06.2026