The place of content ranking algorithms on the AI risk spectrum

Abstract

This article discusses how content ranking algorithms are positioned within the AI Act. It seeks an answer the quesetion whether content ranking algorithms could count as high-risk systems under the Act? It examines the possibility of later inclusion into the list of high-risk systems by Commission action. Then, it compares the envisioned obligations of high-risk systems with those provided by the Digital Services Act. The article puts forward arguments that content ranking algorithms should be subjected to strict ethical requirements due to their impact on the democratic public discourse, and therefore indirectly, democracy.