INTELIGÊNCIA ARTIFICIAL NO PODER JUDICIÁRIO BRASILEIRO

UM PANORAMA AMPLIADO DAS INICIATIVAS NOS TRIBUNAIS DA REGIÃO SUDESTE

Abstract

This article presents the findings of the second stage of an empirical, qualitative, and exploratory research project conducted by the Research Group Observatory on the Advancement of Artificial Intelligence in Jurisdictional Activity and the Effectiveness of Access to Justice, affiliated with IBMEC/RJ. Building upon the first stage, which focused on the courts of the State of Rio de Janeiro - this study broadens the scope to encompass a more comprehensive comparative analysis of generative artificial intelligence use in the Brazilian judiciary, with emphasis on the Regional Labor Courts of the Southeast region, the expansion of the State Court analysis to include the courts of Minas Gerais, Espírito Santo, and São Paulo, and institutional developments within the Federal Courts. The research examines the tools deployed, governance structures, training initiatives for judges and court staff, and compliance with CNJ Resolution Nº. 615/2025. Findings reveal a landscape marked by heterogeneity and asymmetry in both technological maturity and institutional transparency. Significant advances are identified - such as the TRF2 development of ApoIA and its integration into the national Digital Platform of the Brazilian Judiciary, alongside substantial gaps in public disclosure of actual usage data, algorithmic auditability, and standardization of continuing education policies. The study reaffirms the imperative of human oversight, protection of sensitive data, and robust technological governance as essential foundations for a responsible and equitable digital transformation of the Brazilian judiciary.