A key part of the Fourth Industrial Revolution, Artificial Intelligence is set to deeply transform every sector of our economy. At a time when geopolitical tensions and international mistrust are reaching alarming levels, harnessing and mastering the power of such an essential tool for the future has become a top priority for many countries.

Being able to control one’s data and provide a safe digital environment for companies and citizens, aligned with one’s own values and interests, is what is today called “sovereign AI” — or a nation or a region’s ability to develop and produce artificial intelligence using its own workforce, data, infrastructure and business networks. In short, and ideally, this means being in command of the entire AI supply chain, thus reducing the technological dependence to foreign solutions (mostly coming from the U.S. and China) that lack transparency over data treatment.

Jumping on board the sovereign AI train has become of vital strategic importance for Europe and its members, like France, who are rife with talents and innovative solutions. Through an approach focused on trust and excellence, frameworked by a strong data protection policy (the General Data Protection Regulation, GDPR) and various data-focused acts (AI Act, Data Governance Act), the EU has all the cards in hand to build a trustworthy and robust ecosystem, and take control of its digital future.

Tapping into Europe’s breeding ground

“France is renowned for having remarkable digital experts,” says Arnaud Muller, co-founder of Cleyrop, Europe’s first industrial and secured data hub. “Take the Mathematics, Vision and Learning Master of the École Normale Supérieure, for instance: it is considered one of the best degrees to produce the next AI champions.” The country also benefits from a solid scientific and academic core regarding digital technologies, both in the public and private sectors. Two good examples of this fertile ground are Kyutai, the first European non-profit private-initiative laboratory dedicated to open research in artificial intelligence, and Probabl:, the startup founded by the National Institute for Research in Digital Science and Technology (Inria) to respond to a government’s order to develop a sustainable data-science commons ecosystem around scikit-learn. Finally, French company Mistral — which produces open-weight large language models (LLM) — is already seen as a bonafide rival to OpenAI and ChatGPT.

From Belgium’s capital city, Samuel Profumo, President of French Tech Brussels, helps Belgian and French start-ups looking to grow in both countries, as well as within the EU. Everyday, a new digital gem knocks at his door, with the potential to become a trusted link in this big, trustworthy, European digital network. “Here in Belgium, we already have strong digital players like Collibra, our very first unicorn which is building data governance infrastructure for the AI era,” he explains, adding that Europe’s goal is “to build this third digital way, grounded on ethics, on the EU’s democratic values and on a human-first principle.” A choice that could become the region’s greatest asset to compete against the traditional giants like GAFAM and China.

Beyond France’s and Belgium’s borders, the European Union also benefits from a broad range of skills. “When it comes to sustainability, which is also a pillar of the third digital way, the Nordics are very advanced,” notes Arnaud Muller, while highly dynamic clusters are emerging from other countries such as Spain, Portugal or Estonia.

Building bridges and boosting investment

More importantly, all of Europe’s AI actors share a common vision and ambition: that of developing a collaborative, human-centric and regulated industry. What is still missing in this ecosystem is making the connection between the solution producers and their users, between the private and public players. “The uses are what will create the value”, highlights Arnaud Muller. “The market is still fragmented,” adds Samuel Profumo. “We need to build bridges to accelerate and scale up this pivotal digital industry.“ In short: shape the market as a local distribution network for AI.

“We could impose quotas on local solutions,” suggests the Cleyrop co-founder. “A softer and interesting approach would also be to create experimental spaces for companies to get to know the solutions available and break down some trust issues.” For big groups, “it would be more a question of hybridization and mixing local and foreign solutions”, ventures Samuel Profumo.

Of course, investments in the infrastructures and businesses are key to achieve AI sovereignty. “I strongly believe that public order can make the difference and encourage more private investments,” says Arnaud Muller. “Governments need to have a more incentivizing approach and a dynamic turned towards supply rather than demand.” For the President of French Tech Brussels, local strategic plans (like France 2030 and its 2.5 billions euros to support artificial intelligence) are showing the way. However, compared to the U.S. and its numerous billion-worth incentives, such budgets can seem too low for those willing to compete.

While financing must increase and collaboration be pushed further, the EU can rely on its comprehensive AI regulatory framework, which is unique in the world, and thus offers significant competitive advantage. “Europe’s regulation represents a real opportunity for the industry but it should not hold back innovation, so we need to make sure that it does allow players to thrive,” concludes Samuel Profumo.