INTELIGÊNCIA ARTIFICIAL NO PODER JUDICIÁRIO: UMA ANÁLISE NA PERSPECTIVA DO MODELO DE ACEITAÇÃO TECNOLÓGICA (TAM)

Abstract

Artificial Intelligence (AI), with its recent advances in areas once limited to human reasoning, such as adaptive analysis and decision-making, is regarded as a potential solution to the chronic delays of the Brazilian Judiciary. Nevertheless, AI remains subject to criticism, risk perceptions, and limited understanding, which may hinder its acceptance. To explore the potential of this emerging technology, this dissertation applies the Technology Acceptance Model (TAM), proposed by Davis (1989), in order to identify the factors influencing AI acceptance within the Judiciary of Mato Grosso do Sul (PJMS). TAM posits that the attitude toward using a technology (ATT) depends on its perceived ease of use (PEOU) and perceived usefulness (PU). A quantitative approach was employed, using a structured questionnaire administered to 507 respondents, with data analyzed through Structural Equation Modeling. The results confirmed the validity of TAM in the judicial context, showing satisfactory fit indices (CFI and TLI close to 1; RMSEA and SRMR within accepted thresholds). All paths were statistically significant, particularly the effect of PEOU on PU (β = 0.53; 95% CI: 0.512–0.741) and of PU on ATT (β = 0.649; 95% CI: 0.720–1.130). The findings provide the PJMS administration with a relevant situational analysis, underscoring the need to strengthen training programs—identified both as the main barrier and the most frequently suggested priority—alongside other recommendations for the safe and responsible implementation of AI. Keywords: Artificial Intelligence; Judiciary of Mato Grosso do Sul; Technology Acceptance Model (TAM)