Français
AI Action Summit: Is regulation holding back innovation?
Alexandre Zapolsky, co-founder and president of LINAGORA, appeared on France 24 to discuss the future of AI governance and its impact on the global economy, following the AI Action Summit held in Paris on 10 and 11 February 2025. Unlike previous summits, the one in Paris brought together not only governments, but also entrepreneurs and civil society. This diversity illustrates that AI is not just a business issue, but a global societal challenge.
Is investment really the solution? At the summit, Emmanuel Macron announced €109 billion in private investment to support AI in France, while the President of the European Commission, Ursula von der Leyen, spoke of a €200 billion European plan (the European Union is committing €50 billion, and €150 billion from major groups). However, Alexandre Zapolsky questions the size of these sums:
‘Since the emergence of open source AI and models like DeepSeek, we have understood that it is not necessary to raise billions to be competitive.’ It's not a question of money, but of impact. In his view, the issue is not funding, but access to quality datasets. ‘In open source AI, we need access to open databases to train our models. That's where the real challenge lies.’
A few weeks ago, LINAGORA launched LUCIE, a LLM Foundation entirely open source where users thought they would find a ChatGPT. But LUCIE is first and foremost a research project.
‘We launched lucie.chat to enable the OpenLLM community to interact with it, but we weren't expecting such a huge response. Many have been testing it as if it were a final product, whereas it's still in the learning phase.’
The main challenge remains the post-training of the model, a key stage where LUCIE must learn to interact better with humans.
‘In France and Europe, we have the energy, we have a community, and that makes all the difference.’
Europe has the assets to compete with the United States and China in the field of AI, but it still faces several major challenges:
- Clearer, harmonised regulation to enable companies to innovate without excessive administrative barriers.
- Easier access to open source datasets, which are essential for developing high-performance models.
- Investment in skills, because one of the main obstacles to the adoption of AI in businesses remains the lack of trained talent.
Watch the interview HERE
