CEPS expert warns Europe needs AI sovereignty and faster innovation

By AI, Created 3:30 PM UTC, May 21, 2026, /AGP/ – Pierre-Alexandre Balland said on the RegulatingAI Podcast at TechEx in San Jose that Europe cannot rely on foreign AI systems for critical services and must move faster on funding, open-source development and local infrastructure. He argued the region’s biggest gap is investment, not regulation, as AI competition with the U.S. and China accelerates.

Why it matters: - Balland said countries that depend on foreign AI platforms risk losing access to systems that run hospitals, schools and businesses during geopolitical conflict. - Europe’s ability to build competitive AI companies could shape whether the region becomes a user of imported tools or a producer of its own technology stack. - Physical AI and automation could reshape labor markets, especially in advanced economies where repetitive work is easier to replace.

What happened: - Pierre-Alexandre Balland, chief data scientist at the Center for European Policy Studies and visiting professor at Harvard University’s Growth Lab, spoke with Sanjay Puri on the RegulatingAI Podcast at TechEx in San Jose. - Balland said Europe needs stronger AI sovereignty, more open-source models and faster innovation. - Balland argued that Europe’s main challenge is funding, not regulation. - Balland said the European Union’s AI Act had good intentions but took too long to implement.

The details: - Balland said investment in AI startups is much higher in the United States than in Europe. - Balland said Europe’s fragmented market makes it harder to build unified technology champions that can compete globally. - Balland said countries such as France and Germany often compete separately instead of acting as a larger European bloc. - Balland said Europe lacks the pension fund structure needed to channel major capital into AI startups and hyperscalers. - Balland said regulators still struggle to define where many AI rules apply, which makes enforcement difficult. - Balland said the AI Act was designed to limit harmful surveillance and protect people from unsafe AI use. - Balland said open-source AI gives governments and businesses more flexibility because models can run locally instead of relying on outside providers. - Balland said cognitive AI tools are already spreading quickly across the Global South through smartphones and apps like WhatsApp. - Balland said robotics and physical AI could disrupt millions of physical labor jobs. - Balland said developing countries may adopt physical AI more slowly because robotics is expensive and depends on infrastructure and supply chains. - Balland said governments in the U.S., Europe and China still influence the AI industry through regulation, trade controls and national policy.

Between the lines: - Balland’s core argument is that AI policy is now an industrial policy question, not just a safety question. - The emphasis on open-source models suggests Europe may need to compete differently than the U.S. and China if it wants more control over critical infrastructure. - His comments on physical AI point to a second wave of disruption beyond chatbots and software, with labor-market effects likely to become a bigger political issue.

What’s next: - Balland said political courage, investment reform, international cooperation and open-source innovation will determine whether countries build independent and trustworthy AI ecosystems. - Europe’s ability to speed up AI investment and implementation will be a key test of whether its policy goals can translate into global competitiveness.

The bottom line: - Balland’s message was blunt: Europe cannot regulate its way to AI leadership without more capital, faster execution and greater control over the technology it depends on.