Notified Bodies in the AI Act: regulatory intermediaries with a de facto gatekeeping role

This paper examines the significant yet underexplored role of notified bodies within the EU AI Act. These conformity assessment entities, once designated by national authorities and notified to the European Commission, are tasked with determining whether high-risk AI systems meet legal requirements to enter the EU market or put into service. Positioned between state regulators and providers of high-risk AI systems, notified bodies function as de facto gatekeepers, with the authority to allow or deny market access. Drawing on gatekeeping theory and regulatory governance literature, the paper conceptualises notified bodies as regulatory intermediaries who wield gatekeeping power despite not being state actors. The paper analyses their functions, accountability, and overall role, highlighting tensions arising from their quasi-public responsibilities and private nature. The paper argues for stronger safeguards to ensure legitimacy, transparency, and fundamental rights protection