The Impact of AI on Separate and Dissenting Opinions in International Courts: Evidence from the Inter-American Court of Human Rights

Courts around the world are increasingly confronted with the challenges posed by the integration of artificial intelligence (AI) into judicial processes. The incorporation of AI-generated information into the judicial “corpus” of decisions has the potential to affect not only the final judgments of courts, but also the separate and dissenting opinions issued by judges, as well as other institutional outputs such as press releases and summaries of decisions. Despite the growing relevance of this development, little is currently known about how domestic and international courts are using AI in their daily operations or how such use may influence judicial reasoning and legitimacy. This paper seeks to address this gap by examining the role of AI in the production of separate opinions within the Inter-American Court of Human Rights. The central research question guiding this study is: Can we trust AI to collaborate in the production of separate opinions in international courts? It focuses on the introduction of THEMIS, an AI-based program recently implemented by the Court to support judicial work in 2019. Using the Inter-American Court of Human Rights as a case study, this paper analyses dissenting and separate opinions issued both before and after the implementation of the THEMIS system. To this end, more than 120 separate opinions have been collected and systematically examined. The study distinguishes between opinions that were produced before and after the AI system implementation. By comparing these two sets of opinions, the paper aims to identify potential discrepancies in style, frequency, length, reasoning, and substance that may be associated with the use of AI tools. The main objective of the paper is to assess whether the introduction of AI has altered the practice of issuing separate opinions, either by facilitating their production or by influencing their content. Particular attention is paid to whether AI-assisted drafting correlates with an increase in the number of separate and dissenting opinions, and whether such changes raise concerns about coherence, judicial authority, or institutional legitimacy. The analysis therefore goes beyond technical considerations and engages with broader normative questions surrounding trust in judicial decision-making. More broadly, this paper contributes to the emerging literature on AI and trust in judicial institutions. It engages with debates on authority and legitimacy in international courts, particularly in contexts where technological tools may shape legal reasoning. It explores the legal, sociological, and technical challenges faced by the judiciary when adopting AI systems, and how these challenges may affect public and institutional trust in court systems. Finally, the paper contributes to the literature on judicial politics and polarization in constitutional and international courts. While a reasonable level of separate and dissenting opinions can enrich jurisprudence and promote legal dialogue, a dramatic increase in such opinions may undermine legal certainty and institutional polarization. Understanding the role of AI in this dynamic is therefore essential for evaluating its long-term implications for international adjudication.