European firms diversify AI vendors as U.S. access limits bite
U.S. access limits are forcing European firms onto multiple AI vendors, turning model choice into a resilience test for procurement, risk and product teams.

European companies are no longer choosing AI providers on performance alone. At VivaTech in Paris, Siemens, Renault Group, Orange and ChapsVision said they were already mixing U.S., Chinese and European models as access limits and sovereignty worries made single-vendor dependence harder to justify.
The pressure sharpened after the U.S. government ordered Anthropic to suspend access to its Fable 5 and Mythos 5 models for foreign nationals on national security grounds. Siemens said it uses DeepSeek, Alibaba’s Qwen and Nvidia’s Nemotron alongside other U.S. and European models. Cedrik Neike, a Siemens executive, said, “You need flexibility,” and argued that sovereignty should not be confused with autarky. Orange said its infrastructure could run all open-source models, including Chinese ones, if they are self-operated on European infrastructure.

That mix-and-match approach reflects a market that is still thin on homegrown options. Europe’s general-purpose AI provider base is small, with France’s Mistral leading that field and DeepL stronger in narrower use cases. For large companies, that means the answer to AI procurement is increasingly not which model is best, but how much optionality a contract, an architecture or an internal platform preserves if access changes overnight.
The shift has direct implications for Goldman Sachs teams that advise, build or police AI-enabled products. European clients are now likely to press harder on vendor concentration, cross-border access, model portability and what happens if a provider becomes unavailable or politically constrained. That turns procurement into a resilience exercise and pushes new diligence work onto bankers, technologists, operations staff and risk officers who have to assess not just performance, but continuity, data controls and cyber oversight. Spreading workloads across several providers can reduce lock-in, but it also raises cost, complexity and governance demands.
The policy backdrop is moving in the same direction. The European Commission unveiled its European Technological Sovereignty Package on June 3, including the Chips Act 2.0, the Cloud and AI Development Act, the EU Open Source Strategy and a strategic roadmap for digitalisation and AI in energy. The Commission says the bloc relies on non-EU countries for over 80% of key digital products, services, infrastructure and intellectual property, a dependency that is now shaping how European firms buy, deploy and defend AI systems.
This article was produced by Prism’s automated news system from verified source data, official records, and press releases, then run through automated quality and moderation checks before publishing. The system is built and supervised by the people who set the standards it runs under. Read our full AI policy.
Did this article answer your question?

