Integrating Artificial Intelligence in Ukraine’s Courts

War can have far-reaching consequences, impacting everything from economic stability to governance transformations. In Ukraine, the ongoing conflict with Russia has been both a barrier to reform and, paradoxically, a catalyst for modernization in certain areas. One notable example is the country’s drive to integrate artificial intelligence (AI) within its defense, public administration and courts – a bold move aimed at adjusting to challenges brought by war realities, and enhancing efficiency, transparency, and access to justice, even in the face of considerable obstacles.

Despite the ongoing Russian invasion, Ukraine is making significant progress in digitalizing its public sector, including efforts to incorporate AI tools in the judiciary. This offers a unique opportunity to counter – at least to some extent – longstanding issues such as inefficiencies in dealing with cases (by e.g. automating case management), and even corruption within the courts (e.g. by using AI tools to assign cases to judges and, in this way, limiting the possibility of selecting ‘the preferred judge’).

This post examines Ukraine’s recent steps toward AI integration in the courts, highlighting initiatives and plans for the future. While these efforts reflect a growing recognition of AI’s potential, they also reveal limitations. Concerns surrounding AI, such as data security and confidentiality, reliability, transparency, explainability, accountability, fairness, and bias, are just as significant in judicial contexts as they are in other areas. For AI to genuinely benefit Ukraine’s judiciary, the implementation must be transparent, strategically paced, and focused on strengthening—not replacing—the role of human judgment. 

AI in Courts: A Short Note on the Concept

Defining ‘AI’ and distinguishing it from non-AI machine systems remains challenging. Despite efforts by leading thinkers, no universally accepted definition of the term ‘AI’ exists. In general, AI refers to a wide range of systems that mimic aspects of human intelligence, or, in other words, are able to perform tasks that normally require human intelligence. The terms AI and AI system are thus employed to cover various techniques, including machine learning and knowledge-based methods, generative AI, as well as application areas such as computer vision, natural language processing, speech recognition, intelligent decision-making systems, intelligent robotics, and the innovative use of these tools across different fields.

This blog post focuses on AI within courts, including both AI that can be used by courts as public institutions and AI used by judicial assistants and judges when working on a case. In courts, AI can be employed for various tasks and at all stages of proceedings. First, AI can be employed in courts for technical, administrative, and secretarial functions, including case management, document processing, scheduling, transcription of oral arguments or translations. Second, AI may play a role in assisting judges and their legal assistants in providing legal research and drafting support. Lastly, AI can also be used to replace human decision-making (see, e.g., research made by E. Volokh and T. Sourdin), though no country so far seems to have fully entrusted the judge’s role to AI, at least without any intervention from a human judge.

First Steps in Integrating AI in Ukrainian Courts

In Ukraine, efforts to reform the judiciary have often been closely linked with the ongoing war. While initiatives to strengthen the rule of law through judicial changes can be traced back to the onset of the conflict in 2014, the digitalization process in Ukrainian courts began only in 2018, particularly with the presentation of the ‘Unified Judicial Information and Telecommunication System’ (UJITS). This initiative aimed to create a paperless system for managing documents and automatically assigning cases to judges, promoting transparency and open justice. After six years, we can observe that the system has started to automate various aspects of the judicial process, including document management, office tasks, analytics, and communication between courts using electronic data exchange, video conferencing and digital signatures. The subsequent launch of the ‘Electronic Court’ (one of the sub-systems of UJITS) provided participants in legal proceedings with online tools to submit claims, monitor case progress, pay court fees, and even participate in video conferences, etc. While this program did not integrate AI instruments, it laid the groundwork for future improvements.

In September 2020, the Ukrainian Ministry of Justice launched the first AI-enhanced system for judicial use, called Cassandra, which assesses the risks of reoffending among prisoners. Since its launch, updates on use or ongoing success have been limited. Available information indicates that probation officers input data based on specific questions, which then scores to estimate the likelihood of reoffending. It can be noted that risk prediction systems, predicting risks for recidivism or violent crimes, were already deployed in other countries (the most notorious example from Western countries is the US). However, such tools are not that straightforward and raise many concerns (see, e.g. F. Palmiotto, C. McKay, J. LM McDaniel and K Pease). Research has already shown that the use of predictive tools might exaggerate bias and discrimination in police (see, e.g. Fair Trials) and in criminal court activities (see a recent study by Ch. Engel, L. Linhardt and M. Schubert), thus, particular care should be taken, and specialized training should be offered to those using such tools.

That same year, the Cabinet of Ministers approved the ‘Concept for the Development of Artificial Intelligence in Ukraine,’ which included plans to further integrate AI into the sphere of justice. In 2021, the Supreme Council of Justice proposed a pilot project to use AI in first-instance courts for administrative offences. The goal was for an AI system to automate the handling of small offences, such as minor traffic or parking violations, streamlining how laws are applied and helping identify issues in legal interpretations. This initiative coincided with the challenges posed by the COVID-19 pandemic, as judicial authorities worked with the Ministry of Digital Transformation to help ease the burden on the judicial system by exploring AI. However, the implementation of these plans was disrupted by the Russian invasion of Ukraine in February 2022. With the country’s focus shifting toward preserving its sovereignty, many reforms, including those related to AI in the judiciary, were put on hold.

Resumption of AI Integration

With the war continuing, progress has resumed as the justice system recognizes the urgent need to enhance judicial reforms to meet European integration requirements. EU Commission reports on Ukraine reveal that to increase transparency, efficiency, and access to justice, Ukraine should continue its efforts on digitalization of the judiciary. According to the European Commission for the Efficiency of Justice (CEPEJ), Ukraine’s ICT deployment and usage index (measuring digital access to justice; case management and decision support) is way below the Council of Europe’s average.

In 2023, an improved AI-driven tool was introduced to create an open database of Supreme Court decisions, making it easier to find relevant legal information. Initially, using a standard search algorithm, the tool was improved by adding AI that compared legal terms in submitted documents with relevant Supreme Court rulings. Further efforts are expected to focus on integrating AI into other projects, such as the Unified State Register of Courts Decisions of Ukraine – which contains around 120 million court documents and is still used for searching lower courts’ practice. Besides that, the Unified Judicial Information and Telecommunication System, despite launching new subsequent systems, still needs improvement by integrating AI tools. However, it is important to note that such developments require substantial time and financial investment, and the potential for change remains unclear in the economic crisis caused by the war.

The discussion of implementing AI instruments extends beyond individual projects. During an All-Ukrainian Seminar for Judges in April 2024, the Supreme Court discussed how tools like OpenAI’s ChatGPT and Microsoft’s Gemini could assist judges by finding relevant legislation or case law, translating European Court of Human Rights (ECHR) cases, checking court decision drafts for grammar, and even verifying the authenticity of evidence. While these ideas may be quite promising, they raise significant concerns about conformity with judicial values and human rights. While researching publicly available legal acts and case law generally poses no issues, tools that rely on specific case materials or internal documents of courts must be carefully tailored and securely used within the Ukrainian court system. The use of external tools needs to follow clear rules; without them there is a risk of compromising data security for participants in legal proceedings.

Discussions around generative AI in courts have also intensified, especially following a Supreme Court decision on 8 February 2024. In this case, an applicant argued that a definition of “voluntary obligation” generated by ChatGPT should prompt the court to clarify its previous use of the term. The Supreme Court rejected this approach, stating it was an abuse of procedural rights. The court emphasized that AI should not be recognized as “a source of reliably proven scientific information” and using AI-generated information to challenge the court decision could undermine the authority of the Supreme Court and erode public trust in the judiciary. Although this case focused on the applicant’s use of AI rather than the court’s, it highlights the distinct stance of the Supreme Court toward AI. This difference in perspective shows the need for a cohesive framework to guide AI implementation in judicial processes.

Finally, just recently, in September 2024, further development were introduced by amendments to the Code of Judicial Ethics, which now allows judges to use AI technologies in their work. The amendments state that “the use of artificial intelligence technologies by a judge is admissible if it does not affect the independence and impartiality of the judge, does not affect the evaluation of evidence and the decision-making process, and does not violate the requirements of the law”. These changes indicate a strategy to expand the role of AI in the judiciary, despite its limited use in Ukraine so far.

Concluding Remarks

In Ukraine, as in any other democratic country, integration of AI into the court system should align with human rights and judicial values, ensuring that these technologies enhance rather than compromise justice. AI has the potential to increase the efficiency of administrative work in courts and help court staff with technical assignments. However, introducing AI, especially generative AI, into the work of judges should be approached with caution, given its current state of development.

The current situation indicates that Ukraine still stands at the beginning of integrating AI into its court system. Recent challenges, such as the COVID-19 pandemic, the Russian invasion, and the economic crisis, have significantly affected Ukraine’s ability to show substantial progress in this area. Furthermore, the AI integration process appears somewhat chaotic and opaque, with limited information publicly available about the projects and plans for improvement.

Nevertheless, recognition of AI’s potential utility is steadily growing, and some promising developments have emerged even during wartime. We hope that recent legislative changes will lead to potential evident advancements in this field, despite the ongoing military conflict delaying the progress.

Yuliia Moskvytyn and Dr. Agne Limante

Law Institute of the Lithuanian Centre for Social Sciences