Conformity assessment procedures under the AI Act

This article critically analyses conformity assessment procedures for high-risk AI systems under the AI Act. It maps the applicable procedures to different types of AI systems, offering structured guidance for implementation. The article then evaluates the AI Act’s reliance on procedures laid down in other Union harmonisation legislation. It demonstrates how this reliance leads to a risk classification that fails to adequately consider AI-specific risks, and questions the suitability of various procedures for assessing AI systems. Finally, it explores how the assignment of these conformity assessment procedures impacts the requirements that conformity assessment bodies must adhere to. By offering this analysis of the regulatory landscape, this article contributes to a better understanding of the AI Act’s systemic challenges and identifies regulatory blind spots.