In the second in-depth essays on the proposal for a EU Regulation on Artificial Intelligence - published today by Giuffrè Lefebvre Publisher - Prof. Avv. Alessandro del Ninno carries out a practical analysis on the risk-based approach that arises from the proposed IA Regulation (also comparatively with the risk-based approach adopted by the General Data Protection Regulation no. 679/2016).

The examination of the classification of high-risk Artificial Intelligence systems and the start of the description - for these systems - of the practical obligations to those who place them on the market or put them into use follows. In fact, suppliers must "design, implement, document and maintain" (based on the provisions of Article 9 of the proposed IA Regulation) an overall risk assessment system for high-risk AI systems, made up of legal, technical and documentary requirements.