Abstract
In discussions surrounding the application of artificial intelligence in the legal field, the concept of “predictive justice” frequently arises. This refers to computational systems based on machine learning and deep learning that analyze vast amounts of data to anticipate the outcomes and content of future judicial decisions. Among these tools, judge profiling elicits particular interest, as it aims to predict not just how the law will be applied in general but specifically how an individual judge will decide a case. Research into judge profiling in the United States reveals that such systems analyze not only legal texts and judicial precedents but also biographical and professional data about the judge in question. While predictive tools like judge profiling promise enhanced certainty and equality, they also introduce significant challenges. Implementing such systems could lead to issues like the alteration of observed phenomena and the behavior of the judge due to increased scrutiny. This, in turn, could undermine judicial independence and the effectiveness of rights protection. Is France’s decision to criminalize judge profiling practices in 2019 the only viable approach? The balance between the benefits of increased predictability and the risks to judicial independence is a critical concern that warrants careful consideration, also in light of the European Regulation on Artificial Intelligence (“AI Act”).
Presenters
Francesco Maria DamossoPostdoctoral Researcher, Dipartimento di Economia, Ingegneria, Società e Impresa, Università degli Studi della Tuscia, Agrigento, Italy
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
Judge Profiling Tools. Predictive Justice. AI Act
