Europe’s tech dependence exposes limits of Brussels regulation

A US export-control order turned AI sovereignty into an immediate concern for Europe. On June 12, the US Commerce Department’s Bureau of Industry and Security directed Anthropic to cut off its newest frontier models for all foreign nationals, prompting the company to block public access worldwide when it could not segment users by citizenship. Commerce rescinded the order on June 30 after Anthropic negotiated cybersecurity safeguards and testing protocols, while reports also described pressure on OpenAI’s GPT-5.6.

Europe’s vulnerability spans software, chips, cloud, and frontier models. EU policymakers acknowledge that 80% of EU enterprise-software spend goes to US firms, while advanced chips depend on a chain involving Nvidia, TSMC, and ASML. In cloud, the three US hyperscalers hold roughly 70% of the European market, European providers remain in the low teens, and the EU spends an estimated €264bn a year on US tech.

Procedural regulation has delivered limited leverage against entrenched platforms. The Irish Data Protection Commission has yet to collect over €2.8 billion in fines against Meta and nearly €900 million against TikTok, while Apple, Microsoft, and US political pressure have complicated enforcement of EU rules. A proposed path combines building European capacity through the EU Cloud and AI Development Act, procurement preferences, talent attraction, and investment, then constraining systemic firms through structural remedies and market-shaping standards rather than slow fines and compliance processes.