Legal and non-legal norms in AI systems for decision making in public administration
Talk by Søren Stig Andersen, PhD, legal researcher (UCPH).
The use of semi-automated and fully automated administrative decision-making systems may challenge legal compliance in a deep sense. While some legal requirements are widely known and relatively straightforward to implement in automated systems, deeper-lying legal principles and values may pose bigger challenges. This is especially the case when individuals with professional qualifications outside the legal field significantly influence the implementation of such systems. On the other hand, such involvement may increase the impact of extra-legal norms originating from other professions.
Drawing on insights from a study published in 2025 that combined qualitative empirical research and doctrinal law, Søren will present how the implementation of AI systems for decision making in public administration can impact administrative practice.
The background study is available here
Registration
A sandwich will be served to those who sign up.