Digital

Allan Deneuville_VF
π Society π Digital

Somewhere between a tool for citizens and a surveillance weapon, what exactly is OSINT?

With Allan Deneuville, Lecturer at Université Bordeaux Montaigne and Researcher at GEODE (Geopolitics of the Datasphere)

On April 14th, 2026
4 min reading time
π Digital π Industry

European AI gigafactories: the true, the false and the uncertain

With Nicoleta Kyosovska, Research Assistant at the Centre for European Policy Studies (CEPS)

On March 24th, 2026
4 min reading time

The call for tenders requires AI factories to have plans for collaboration with one another, yet, for the time being, public information on this subject is limited.

π Digital π Society

How is generative AI changing modern education?

With Michel Barabel, Associate Professor at Sciences Po and Researcher at Université Paris-Est Créteil

On March 18th, 2026
5 min reading time

Current limits of generative AI in education are less about technical performance than about effects on learning trajectories.

π Digital π Geopolitics

Electric dependency: what strategies for the EU?

With Gilles Lepesant, CNRS Research Director in Geography and Lecturer at Sciences Po Paris

On March 18th, 2026
4 min reading time

Coveted by the industrial and defence sectors, rare earth elements have become critical components of sovereignty.

π Society π Digital

The addictive design of social media under fire from regulators 

With Lê Nguyên Hoang, PhD in Mathematics (Polytechnique Montréal), Co-Founder and Executive Director of Tournesol.app

On March 11th, 2026
5 min reading time

Beyond condemning TikTok, YouTube, and Meta, it seems more reasonable to invest in developing alternatives that comply with democratic standards.

π Digital

What are the markers of trust for generative AI?

With Christophe Gaie, Deputy Director of the Digital Operations Division at the AP-HP, Jean Langlois-Berthelot, Doctor of Applied Mathematics, Former Researcher at Ecole Polytechnique (IP Paris)

On February 4th, 2026
4 min reading time

Integrating trust markers into AI could enable users to assess the relevance and credibility of the model according to their needs.