Institutional Openness and the Limits of Predictive Legal Output

This project examines the limits of predictive legal output in public-law decision-making. It asks whether AI-assisted legal systems can do more than retrieve authorities, summarize doctrine, predict outcomes, or generate plausible legal reasons. The central concern is that legal validity cannot be reduced to output production, because public law depends on institutional processes through which decisions remain open to challenge, justification, correction, and review. The project distinguishes b