In the aisles of VivaTech, artificial intelligence demonstrations come one after another at a breakneck pace. Yet behind the spectacular announcements and promises of technological breakthroughs, a more fundamental question remains: Why does Europe still struggle to turn innovation into sustainable productivity gains?
It was around this question that Xavier Jaravel, a professor of economics at the London School of Economics and president of the Economic Analysis Council, and Margot Dufourcq, a partner at Serena, a European venture capital firm specializing in deep tech and artificial intelligence, came together.
Their conclusion is clear-cut. For three decades, Europe has been falling further and further behind the United States in terms of productivity. According to Xavier Jaravel, if the continent had followed the same trajectory as the U.S. economy, GDP per capita would be about 40% higher today.
Contrary to a widely held belief, this gap cannot be explained primarily by the absence of tech giants comparable to Google, Amazon, or Microsoft. These companies account for only a fraction of the observed gap. The real difference lies elsewhere: in the ability to rapidly disseminate innovations throughout the entire economic fabric.
In other words, the problem in Europe is not so much invention as adoption.
History is repeating itself today with generative artificial intelligence. The United States has significantly higher adoption rates in the workplace. While about 13% of U.S. workers make extensive use of generative AI in their daily work, that figure is estimated to be around 7% in Europe.
This difference may seem minor. It is not.
Artificial intelligence is not just a technology. It acts as an organizational catalyst. Companies that integrate it effectively are rethinking their processes, automating certain tasks, improving their analytical capabilities, and shortening their decision-making cycles.
According to the studies cited during the discussion, organizations that are the most advanced in adopting AI do not eliminate more jobs than others. In fact, they tend to retain their workforce more effectively, including in low-skilled positions.
The real risk, then, is not being replaced by AI.
The risk is being replaced by a company that makes better use of AI.
This nuance profoundly changes the nature of the debate.
For several decades, Europe has approached major technological shifts with caution. The advent of cloud computing is a telling example. According to Serena, nearly 75% of U.S. small and medium-sized businesses have migrated their infrastructure to the cloud. In Europe, the rate remains below 50%.
Yet this transition has become a prerequisite for artificial intelligence. A cloud architecture provides structured data, documented APIs, and infrastructure that is flexible enough to quickly deploy new tools.
In many companies, AI is now being integrated into existing software. Office suites, CRM systems, and collaborative platforms are gradually incorporating generative assistance features. Adoption no longer necessarily involves creating complex projects, but rather enabling features that are already available.
This reality explains why the issue of technology is now inextricably linked to the issue of management.
Margot Dufourcq describes technology adoption as an organizational muscle. The companies that grow the fastest are often those whose leaders themselves have a practical understanding of artificial intelligence.
A culture of experimentation then becomes a competitive advantage.
Identifying use cases, deciding between in-house development and existing solutions, training teams, and measuring actual gains: these are all skills that fall more within the realm of management than technology.
In this vein, the issue of sovereignty must not become an excuse for inaction.
The speakers advocate a hybrid approach. Certain strategic infrastructures do indeed warrant European control. But waiting for a fully autonomous ecosystem to emerge before taking action would only cause us to fall further behind.
Europe already has credible players. Mistral AI in France is the most visible example. More broadly, the continent continues to produce top-tier talent, who are often recruited by the world’s largest technology companies.
The question is no longer whether Europe has the necessary expertise.
The question is whether she will be able to turn them into widespread adoption.
Because tomorrow’s productivity will not depend solely on the most powerful models or the most sophisticated infrastructure.
It will depend above all on how quickly organizations are willing to reinvent themselves.
Artificial intelligence is not a standalone technological revolution. It acts as a catalyst. It reveals which companies are capable of learning quickly and which remain trapped in their old ways.
In this race, time may be the scarcest resource.
And the most strategic one.
Cette publication est également disponible en :
