On Thursday, American chipmaker Nvidia announced its acquisition of Hugging Face, the AI library and community hub, for $13 billion (€11.2 billion). The price tag is staggering for a platform that began as a teen-focused chat app but has since evolved into one of the most important repositories for open-source and open-weight AI models.
Although the deal is primarily a US affair, Hugging Face's co-founders are French, and roughly half of its employees are based in Paris. The company has positioned itself as a cornerstone of the open-source AI ecosystem in Europe, and its acquisition by Nvidia is likely to draw scrutiny from regulators in both Paris and Brussels, particularly in light of the EU's AI Act and broader digital sovereignty ambitions.
Open vs. closed: the stakes for Europe
Nvidia has become the world's most valuable publicly traded company, riding a wave of demand for its GPUs in AI data centres. In recent months, the company has embraced open AI models, offering dozens for free on its platforms and advocating for competition between closed-source and open-source providers.
For European innovators who lack the resources to train and launch models on the scale of OpenAI or Anthropic, Hugging Face's platform could be a vital tool. As Hugging Face co-founder Clément Delangue told Le Monde in 2024, “For many of the problems that exist in AI, open-source can be a solution. Because the proprietary models are black boxes that make it difficult to analyse their bias or sources of truth.”
The EU's approach to AI has been to set rules and policies that serve as a global blueprint, often favouring domestic industries and continental goals. But innovation is happening across the entire stack, not just among European LLM makers like Mistral or DeepL, which are covered by the EU AI Act. Dutch company ASML, for instance, is the largest supplier of semiconductor manufacturing equipment worldwide, while ElevenLabs, founded by Polish researchers, has become a leader in speech and voice processing, integrated into thousands of AI products.
The push for sovereignty
Beyond commercial innovators, open models can be customised for government agencies, hosted locally and trained on sensitive data. Vendor lock-in is a real concern for EU regulators, as Nvidia's acquisition of Hugging Face could exacerbate or alleviate these worries. Open models could provide genuine competition and reduce reliance on larger tech firms that store data outside the EU.
Building EU data centres is a major goal of European regulators, but the software running inside those centres matters just as much. Open-source models could help de-risk European systems from dependence on American or Chinese infrastructure, though this would require significant expertise within government agencies.
For everyday users, consumers, and small companies, the most exciting aspect of open models is their accessibility at little or no cost. This is a boon to researchers and practitioners in the arts and sciences who may not have the budgets of large corporations.
“Open science and open source AI distribute economic gains by enabling hundreds of thousands of small companies and startups to build with AI. It fosters innovation, and fair competition between all,” Delangue testified before a US House Committee in 2023.
Whether the acquisition will ultimately strengthen Europe's AI sovereignty depends on how regulators respond and how Nvidia manages Hugging Face's open ethos. The deal could either cement Nvidia's dominance or, if handled well, provide European developers with the tools to build independent AI solutions. As Nvidia's revenue surges on AI demand, the company's moves will be watched closely on both sides of the Atlantic.
