This article is taken from the monthly magazine Sciences et Avenir n°949, dated March 2026.
Transparency. Explainability. Reliability. These concepts are repeated like a mantra in the field of artificial intelligence (AI) based on data learning. But they become imperative in the case of medical applications of this AI. This is precisely the aim of the four-year research partnership established between Inria (the French National Institute for Research in Digital Science and Technology) and Doctolib, announced at the end of November 2025.

