Artificial Intelligence About Itself: An Empirical Analysis of Its Readiness to Participate in the Judicial Stage of Criminal Proceedings in the Russian Federation

In the conducted study, open AI models acted as respondents for the first time, answering questions about the possibility and prospects of their participation in the judicial stage of criminal proceedings in the Russian Federation. Unlike a number of studies on how people would like to see or imagine the AI use in the legal process, the artificial intelligence self‑interview method was actually used here, which allows us to empirically assess the self‑assessment of modern AI models regarding legal barriers, technical capabilities and their understanding of key legal categories. Based on the analysis of the responses received, both general and different AI judgments about its role in criminal proceedings were identified, as well as limitations related to its lack of internal beliefs and moral qualities. The few foreign experiences of integrating AI into the justice system — from the risk assessment of offenders in the United States to the “smart courts” of China and experiments in the automation of justice in Estonia — is considered for the purpose of possible implementation in Russian legislation. The article analyzes the AI interpretation of such concepts as legality, validity and fairness of a court decision, and examines its view on the process of proof and consideration of a criminal case on the merits.