Artificial Intelligence and Judicial State Capacity

This project evaluates whether an AI-enabled courtroom support system can strengthen judicial capacity by reducing administrative burden on judges and improving case processing in Indian courts. The intervention deploys AI-based tools that automate documentation, reducing reliance on handwritten court processes in a context where shortages of stenographers often create bottlenecks. This allows judges to reallocate time toward hearings, research, judgment writing, and case disposal. The study falls within J-PAL’s Political Economy and Governance sector.

The project is implemented in partnership with Adalat AI, a legal-tech organization operating in over 4000 courtrooms across nine Indian states. Building on a preliminary pilot conducted in Odisha involving 80 courtrooms randomly assigned to treatment and control, the study will implement a state-wide randomized evaluation in an Indian state.

The evaluation will measure impacts on judicial productivity (case disposals, adjournments, judgment output) and judgment quality. Primary data will be collected through streamlined surveys, and AI-based text analysis will be used to assess the structure, clarity, and legal reasoning of judgments.

By rigorously evaluating an AI intervention in courts, the project generates novel evidence on whether AI can enhance judicial state capacity and improve access to justice.