Policy Modeling and Reasoning in Sociotechnical Systems (Dagstuhl Seminar 25271)

Abstract

This report documents the program and the outcomes of Dagstuhl Seminar 25271 "Policy Modeling and Reasoning in Sociotechnical Systems". This seminar brought together researchers from academia and industry who are interested in studying the intersection between computer science, philosophy, logic, ethics, and law to discuss policy modelling and reasoning in a world where computers and humans need to work together. After lightning talks, two invited talks, and an open space topic gathering activity, we settled on four topics for deeper discussion in working groups, interspersed by primer talks from the various communities. The four topics were: 1) Concepts: What are the underlying aspects of this interdisciplinary field, and can they be defined consistently? 2) Agentic AI: How can we enable agents to interact and reason with human users through large language models? 3) Standardisation: How can we facilitate data sharing and compliance in international work with competing business interests? 4) Coevolution: How can we make sure that sociotechnical systems evolve with the societies they operate in? This report provides the abstracts of the talks, including participants' lightning talks, the two invited talks, and four primers, along with short reports from each working group detailing their discussions, including challenges and future opportunities.