Seminar: Algorithmic Explanations in the Field
- Datum: 11.02.2025
- Uhrzeit: 14:30 - 16:00
- Vortragende(r): Arian Henning
- Ort: Max Planck Institute for Human Development, Lentzeallee 94, 14195 Berlin
- Raum: 316
- Gastgeber: Center for Humans and Machines
- Rubrik: Gesprächs- und Diskussionsformate, Vorträge

The increasing use of algorithms in legal decision-making has led to calls for "meaningful explanations" for affected individuals, particularly when decisions are made by opaque systems such as machine learning or AI. While legal and social science scholars have outlined criteria for legally and ethically acceptable explanations, the field of explainable AI (XAI) has primarily focused on assisting developers in debugging rather than ensuring that decision subjects understand the outcomes. This project seeks to bridge these two perspectives by developing a framework to organize different methods of explaining algorithmic decisions. First, we categorize explanations based on abstract factors such as locality, selectivity, and contrastiveness. Second, we conduct a controlled field experiment to assess how different types of explanations influence individuals’ understanding and acceptance of algorithmic decisions, as well as their likelihood of appealing those decisions. The findings aim to provide insights into which explanations are most effective and ethically appropriate in practice.
Arian Henning is a Ph.D. candidate in law at the Max Planck Institute for Research on Collective Goods in Bonn, Germany. With a background in law, he also brings expertise in experimental behavioral economics and computer science. He has been a visiting researcher at ETH Zürich’s Center for Law and Economics (Spring 2024) and is currently a visiting researcher at the Oxford Internet Institute. His research focuses on the behaviorally informed design of smart institutions, exploring the intersection of public decision-making and explainable AI (XAI).