On January 18, 2024, the SCOR Foundation held a webinar titled: “Artificial Intelligence and Internal models: the case of banking". The speaker was Christophe Hurlin, Professor of Economics at the University of Orléans, Director of the Laboratoire d'Économie d'Orléans (LEO) and member of the Scientific Council of the French supervisory authority (ACPR). Christophe Hurlin proposed a theoretical and practical reflection on the use of machine learning methods in the context of the internal ratings-based (IRB) approach to bank capital requirements. While machine learning is still rarely used in the regulatory field (IRB, IFRS 9, stress tests), recent discussions initiated by the supervisory authorities suggest that this may soon change. This subject is crucial given the growing concerns about the potential financial instability caused by the use of opaque models. Christophe Hurlin questioned, among other things, the performance of AI models versus classical internal models and the trade-off between performance and interpretability. He also presented an initial tentative conclusion concerning the governance of these models.