Artificial Intelligence in the Judicial System of Ukraine: International Experience and Ukrainian Practice

Abstract

This article examines the current state and prospects of using artificial intelligence (AI) technologies in Ukraine's judicial system. It analyses the international and national regulatory and ethical benchmarks for the use of AI in judicial proceedings, in particular the provisions of Opinion No. 26 (2023) of the Consultative Council of European Judges, the European Ethical Charter on the Use of Artificial Intelligence in Judicial Systems and their Environment, and the CEPEJ Guidelines on the use of generative artificial intelligence for courts. The article also outlines Ukrainian organisational approaches, notably the Regulation on the Use of Artificial Intelligence Technologies by Employees of the Office of the Supreme Court.

The main focus is a review of case law on parties' use of generative AI output, references to fictitious court decisions and legal positions, and AI use in drafting judicial acts. The article argues that AI output has no independent evidentiary value and cannot replace the judge’s assessment of evidence, legal analysis or the reasoning of a court decision. It identifies the key risks of using AI: legal ‘hallucinations’, algorithmic bias, opacity, breaches of confidentiality, loss of professional autonomy, and unequal access to technology.

Effective integration of artificial intelligence technologies into Ukraine’s judicial system requires not only the formal enshrinement of relevant rules, but also the development of a coherent institutional culture of responsible use. The introduction of AI must be accompanied by uniform methodological guidelines for courts and participants in the judicial process, improved systems for verifying the authenticity of electronic sources, and mechanisms for verifying results obtained using generative models. Ensuring the digital literacy of judges, court staff, lawyers and other participants in the administration of justice is of particular importance, as it is precisely their level of awareness that determines the limits of the safe and ethical use of such technologies. At the same time, it is emphasised that the digitalisation of the justice system must not lead to the devaluation of the human element, a decline in trust in the judiciary, or the erosion of guarantees of a fair trial.

The article concludes that the acceptable model for Ukraine is the assistive use of AI, under which technology supports administrative, analytical and information processes but does not take over the functions of the judge. It proposes ways of improving the legal framework: statutory definition of permitted and prohibited uses of AI, the introduction of standards for disclosing its use, and the provision of algorithmic audits, professional training for users and safeguards ensuring effective human oversight.