Open Source: The key to achieve Europe's digital sovereignty in the AI era

Abel Samot — February 13, 2025

The AI race is on, and the stakes could not be higher. The United States and China are pouring hundreds of billions of dollars into artificial intelligence, reshaping industries, militaries, and entire economies. Europe, meanwhile, finds itself at a crossroads. It has world-class researchers, a strong regulatory framework, and a deep tradition of open collaboration, but it is dramatically outspent and risks being left behind. The question is no longer whether Europe should act, but how.

The answer lies in open source. History has shown, again and again, that open approaches defeat proprietary ones. From operating systems to databases, from smartphones to web browsers, openness wins in the long run. Europe has a unique opportunity to leverage this dynamic in AI, turning its apparent weakness into a strategic advantage. But it must act now, and it must fund the builders who are making it happen.

We are behind and lack the resources to face tomorrow's AI revolutions

The funding gap between Europe and its competitors is staggering. In January 2025, the US announced Project Stargate, a $500 billion private-sector investment into AI infrastructure, backed by OpenAI, SoftBank, Oracle, and others. This came on top of the billions already flowing into American AI companies: OpenAI alone has raised over $13 billion from Microsoft, Anthropic has secured more than $7 billion, and xAI raised $6 billion in a single round.

China is not far behind. The Chinese government has made AI a core pillar of its national strategy, with state-backed funds and corporate giants like Baidu, Alibaba, and Tencent deploying tens of billions into AI research and infrastructure. DeepSeek, a Chinese lab, recently demonstrated that massive performance can be achieved at a fraction of the cost with open approaches, sending shockwaves through the industry.

Europe's response has been comparatively modest. France announced a 109 billion euro AI plan, but the vast majority of that is private investment commitments, not direct government spending. The EU's total public AI investment remains a fraction of what the US and China deploy. European AI startups raised roughly $5 billion in 2024, compared to over $70 billion in the US. The talent pipeline is strong, with institutions like INRIA, ETH Zurich, and the Max Planck Institute producing world-class researchers, but many of them leave for better-funded positions in American companies.

We simply cannot compete dollar-for-dollar. Trying to outspend the US and China on proprietary AI is a losing strategy. Europe needs a different playbook, one that maximizes the impact of every euro invested. Open source is that playbook.

In history, open source always wins

The pattern is remarkably consistent. Whenever a proprietary technology dominates an industry, an open alternative eventually emerges, gains critical mass, and takes over. It has happened in every major technology category.

Operating Systems

In the 1990s, Microsoft Windows and proprietary Unix variants dominated computing. Linux, created by Linus Torvalds in 1991 as an open-source project, was dismissed as a hobbyist toy. Today, Linux runs over 90% of the world's servers, powers all of the top 500 supercomputers, and is the foundation of Android, the world's most popular mobile operating system. The proprietary Unix market has essentially vanished, and even Microsoft now embraces Linux in its Azure cloud platform.

Databases

Oracle dominated the database market for decades with its proprietary relational database. Then came MySQL, PostgreSQL, and later MongoDB, all open source. Today, open-source databases are the default choice for new applications. PostgreSQL alone is used by companies from Apple to Instagram. Oracle still generates significant revenue from legacy customers, but the future of databases is overwhelmingly open source.

Smartphones

When Apple launched the iPhone in 2007, it seemed like proprietary platforms would dominate mobile forever. Then Google released Android as an open-source operating system in 2008. Today, Android holds over 70% of the global smartphone market. The open approach allowed thousands of manufacturers to adopt it, creating an ecosystem that no single proprietary platform could match.

Web Browsers

Microsoft Internet Explorer held over 95% of the browser market in 2003. The open-source Mozilla Firefox broke that monopoly, and later Google Chrome, built on the open-source Chromium engine, took the lead. Today, virtually every major browser except Safari is built on open-source foundations. The proprietary browser era is over.

The lesson is clear: open source creates a gravitational pull. It attracts developers, reduces costs, accelerates innovation, and builds trust. Proprietary solutions may lead initially, but openness wins over time because it aligns incentives across the entire ecosystem.

Open Source Strengthens European Sovereignty

Technological Independence

When Europe relies on proprietary American or Chinese AI systems, it places its digital future in foreign hands. Every API call to GPT-4, every dataset processed on a US hyperscaler's cloud, every AI model whose weights are hidden behind a corporate wall represents a dependency that can be leveraged geopolitically. We have already seen how quickly technology access can be weaponized through export controls, sanctions, and policy shifts.

Open source breaks this dependency. When the code and the model weights are open, any organization can run them on local infrastructure, modify them for local needs, and ensure continuity regardless of geopolitical shifts. No single company or government can pull the plug.

Security and Transparency

With proprietary AI systems, you have no idea what happens inside the black box. You cannot audit the code for backdoors, biases, or vulnerabilities. You must trust the vendor's claims about safety and privacy.

Open source changes the equation fundamentally. The code is visible, auditable, and verifiable. Security researchers worldwide can inspect it, identify vulnerabilities, and propose fixes. According to industry surveys, 89% of technology professionals believe open-source software is as secure or more secure than proprietary alternatives, precisely because of this transparency.

For European governments handling sensitive data, from healthcare to defense, this transparency is not a luxury. It is a requirement.

Economic Ecosystem

Open source does not just benefit the companies that create it. It generates an entire economic ecosystem of integrators, consultants, contributors, and derivative products. When VLC, the open-source media player created in France, became the most downloaded software in history, it did not just serve users. It spawned an ecosystem of media technology expertise that benefited French and European companies across the industry.

As Jean-Baptiste Kempf, the founder of VLC, has noted, open source creates value that extends far beyond the original project. It trains developers, establishes standards, and builds communities that become the foundation for future innovation. Every euro invested in open source generates returns that multiply across the economy.

Open Source Is Winning the AI Race

The dynamics that allowed open source to dominate operating systems, databases, and browsers are now playing out in AI, and at an even faster pace.

Transparency

Open-source AI models can be inspected, understood, and trusted. In an era where AI is making decisions about healthcare, justice, and finance, the ability to audit a model's behavior is not just desirable. It is essential. The EU AI Act explicitly recognizes this, providing lighter regulatory treatment for open-source AI projects.

Lower Cost

Running open-source models on your own infrastructure can be dramatically cheaper than paying per-token fees to proprietary providers. DeepSeek's R1 model demonstrated that open approaches can match the performance of models that cost orders of magnitude more to develop. For European companies operating on tighter budgets than their American counterparts, this cost advantage is critical.

Faster Innovation

When thousands of researchers and developers worldwide can build on the same foundation, innovation accelerates exponentially. The pace of improvement in open-source AI models has been remarkable. According to the Stanford Institute for Human-Centered Artificial Intelligence, 65.7% of newly released AI foundation models in 2023 were open, up from 44% the year before. The center of gravity in AI research is shifting toward openness.

Data Privacy

Open-source models can be deployed on-premise, ensuring that sensitive data never leaves the organization's infrastructure. For European companies subject to GDPR and for governments handling classified information, this is a decisive advantage over cloud-based proprietary alternatives that process data on foreign servers.

Customization

Open-source models can be fine-tuned for specific languages, industries, and use cases. A European healthcare company can adapt an open model to understand medical terminology in French, German, or Polish. A defense contractor can customize a model for classified applications. This flexibility is impossible with closed APIs.

A leaked internal Google memo captured the situation bluntly: "We have no moat... Open source models are faster, more customizable, more private, and pound-for-pound more capable." That assessment has only become more true since it was written. The open-source AI ecosystem is growing faster than any single company can innovate alone.

Europe has the elements to lead in Open Source but lacks funding

The irony of Europe's position is that it already possesses many of the ingredients needed to lead the open-source AI revolution. What it lacks is the funding and coordination to bring them together at scale.

European Open Source Successes

Europe has produced some of the most important open-source projects in history. Linux was created by a Finnish developer. The World Wide Web was invented at CERN. VLC came from France. MariaDB from Finland. Blender from the Netherlands. Nextcloud from Germany. LibreOffice from Germany. Mistral, now one of the most prominent AI companies in the world, was founded in Paris. Hugging Face, the platform that has become the GitHub of AI, was started by French founders. Scikit-Learn, the most widely used machine learning library with over 2.2 billion downloads, was developed by INRIA in France.

Europe does not lack talent or vision. It lacks capital at scale.

France Leading in AI

France has emerged as Europe's clear leader in AI. The country hosts Mistral, Hugging Face, and a vibrant ecosystem of AI startups. French researchers are among the most cited in the field. The government has made AI a national priority, with President Macron personally championing investment in the sector.

As Mark Zuckerberg and Spotify CEO Daniel Ek wrote in a joint letter: "Europe has the talent and the values to lead in AI, but it needs to move faster and fund more boldly." They are right. The window of opportunity is open, but it will not stay open forever.

European Infrastructure, Education, and Regulation

Europe has several structural advantages that should not be underestimated. Its university system produces excellent engineers and researchers. The regulatory environment, often criticized as burdensome, actually creates a framework of trust that can be a competitive advantage for open-source solutions that meet high standards. European cloud infrastructure is growing, with providers like OVHcloud, Scaleway, and Hetzner offering sovereign alternatives to American hyperscalers.

The EU AI Act, for all its complexity, sends a clear signal: Europe wants AI that is transparent, auditable, and trustworthy. Open source is the natural answer to those requirements.

The Funding Gap

Yet despite these advantages, European open-source companies struggle to raise the capital they need to compete globally. The venture capital ecosystem in Europe is growing but remains a fraction of the US market. European funds are often smaller, more risk-averse, and less experienced with open-source business models.

Michael Jackson, a partner at an early-stage fund, has pointed out that Europe consistently produces world-class open-source technology but then watches it get acquired or outcompeted by better-funded American rivals. The pattern repeats: European innovation, American monetization.

Yann LeCun, Meta's chief AI scientist and a Frenchman, has been one of the most vocal advocates for open-source AI, arguing that open models are essential for competition, safety, and innovation. Yet much of the funding for open AI research still flows through American companies.

Andre Zayarni, co-founder of Qdrant, a leading open-source vector database company based in Berlin, has spoken about the difficulty of raising growth-stage capital in Europe for open-source infrastructure. The technology is world-class, the traction is strong, but the funding ecosystem does not yet match the opportunity.

Anh-Tho Chuong, co-founder of Lago, an open-source billing platform based in Paris, has described a similar challenge: building a category-defining open-source product in Europe while competing against proprietary alternatives backed by American venture capital at scale.

The stories are consistent across the European open-source landscape. The technology is there. The talent is there. The market opportunity is there. What is missing is dedicated, conviction-driven capital that understands open source and is willing to back European founders at the scale needed to build global companies.

Conclusion

Europe faces a choice. It can continue to underinvest in its open-source ecosystem and watch its best talent and technology migrate to the US, or it can recognize open source for what it is: the most powerful strategic lever available to achieve digital sovereignty in the AI era.

The historical evidence is overwhelming. Open source wins. It wins in operating systems, databases, smartphones, browsers, and it is winning in AI. Europe has the talent, the values, and the regulatory framework to lead this movement. What it needs now is the funding and the conviction to back its builders.

Every euro invested in European open-source AI is not just an investment in a company. It is an investment in sovereignty, in transparency, in an economic ecosystem that benefits the entire continent. The tools to compete exist. The founders are ready. The market is enormous. The only question is whether Europe will have the courage to fund its own future.

The next chapter of the AI revolution will be written in open source. Europe must decide whether it will be an author or a reader.