Explainable artificial intelligence for the moderation of contractual penalties by the court

Abstract

The court’s authority to moderate contractual penalties is a key tool for ensuring fairness in contractual relationships, and its application depends on the interpretation of the vague legal concept of an “excessively high” penalty. The current emergence of artificial intelligence in the legal field offers opportunities for analyzing extensive case law, but it faces the “black box” problem, which conflicts with the requirement to justify judicial decisions. In this paper, we examine the potential of explainable artificial intelligence in objectifying the criteria for moderating fines in the context of Slovak law. Based on an analysis of legal doctrine and a practical examination of the technical capabilities of explainable artificial intelligence, this paper assesses whether it is possible to use algorithms to provide legally relevant justification for the mitigation of contractual penalties. In this paper, we conclude that the most suitable model is the use of artificial intelligence as an assistant in a hybrid system, where the final assessment of proportionality remains within the competence of a human judge.