Artificial intelligence and judicial discretion: a comparative legal analysis of the limits of permissible use

• Yuri V. Onosov

  Yuri V. Onosov. Ural Federal University named after the first President of Russia B. N. Yeltsin. Ekaterinburg. Russia

The article examines the problem of the influence of artificial intelligence on judicial discretion. It is substantiated that the issue of using artificial intelligence in judicial activity cannot be reduced merely to the technical modernization of justice. A more significant issue is the limits of ermissible influence of digital systems on the mechanism of judicial law enforcement, including the establishment of factual circumstances, the assessment of evidence, and the formulation of the reasoning part of a judicial act. Based on the analysis of Russian practice, the Chinese model of judicial artificial intelligence, State v. Loomis, the approaches of the European Union, CEPEJ, and the guidelines of the Hong Kong Judiciary, the main limits of the permissible use of artificial intelligence are identified. The author concludes that contemporary judicial practice does not provide grounds for recognizing artificial intelligence as an independent subject capable of resolving a legal case. Rather, it is a matter of algorithmically mediated judicial discretion, in which a judicial act remains the result of an independent intellectual and volitional choice of a judge, while the factual and legal premises of such a choice may be preliminarily formed or structured by a digital tool. The scientific
novelty of the study lies in identifying three contradictions: between efficiency and individualization, between predictability and the independence of judicial inner conviction, and between the formal responsibility of the judge and the actual influence of the algorithm..
Keywords: artificial intelligence, judicial discretion, judicial conviction, reasoning of a judgment, justice, algorithmization, digitalization, limits of adjudication, judicial responsibility, comparative law