When AI goes to court: does technology enhance the decarbonization effect of environmental judiciary? Quasi-experimental evidence from China

As nations pursue ambitious decarbonization targets, questions arise regarding whether technological innovation can amplify the effectiveness of environmental institutions. This study examines how artificial intelligence (AI) integration transforms environmental judiciary capacity to reduce carbon emissions, leveraging China’s establishment of specialized environmental courts beginning in 2007 as a quasi-natural experiment. Employing a triple-differences identification strategy that combines the staggered rollout of environmental courts with spatial and temporal variation in AI patent intensity, we find that AI adoption in cities with environmental courts reduces carbon emissions by 2.65% relative to control groups, with this effect statistically significant at the one percent level. Heterogeneity analysis reveals that this decarbonization impact is substantially stronger in cities with higher AI application maturity, greater urban ecological resilience, and higher industrial carbon dependency, while regional disparities do not significantly moderate the effect. Mechanism analysis, drawing on semi-structured interviews with government officials and quantitative validation, demonstrates that AI enhances judicial performance through improved evidence processing capabilities, increased legal consistency, reduced litigation barriers, and strengthened enforcement monitoring. These findings suggest that digital transformation of environmental adjudication represents a viable pathway for climate governance.