Between Humans and Machines: Judicial Interpretation of the Automated Decision-Making Practices in the EU

Dr Biber will discuss the following chapter: In the EU, judges have recently provided more precise interpretations of automated decision-making (ADM) practices through their judicial rulings. Prominent instances of ADM practices such as fraud detection, teacher placement, credit scoring, and dismissing workers offer valuable insights into how judges interpret these practices. This chapter aims to systematize these practices and explores the role of judicial interpretation in defining these activities and involvement of humans in decision-making processes. The chapter is divided into three sections focusing on machines, humans, and courts. It begins by exploring the concrete uses of ADM in the EU (‘machines’). From there, it delves into how these systems are currently being utilized by taking into account two official reports published by the EU institutions. The research then identifies the socio-technical quality of ADM practices and argues how this quality necessitates meaningful human participation in decision-making processes. The chapter then examines the human-centric provisions of the relevant EU legal instruments surrounding ADM systems and targeting human participation (‘humans’). Finally, it examines four judicial cases which surfaced public and private contexts in the Netherlands, Italy, and Germany (‘courts’). In conclusion, the chapter identifies three key dimensions of judicial interpretation regarding ADM practices: (i) epistemic; (ii) substantial, encompassing socio-technical and legal dimensions; and (iii) methodological. It argues that these dynamics prove the pivotal role of judicial interpretation in comprehending the technical aspects of automation and ensuring meaningful human participation in decision-making processes.

We will be meeting at 11.45am for a 12pm (noon) start. A light sandwich lunch will be provided.