From performance to human-AI coproduction: Reconceptualizing justice administration in the age of artificial intelligence

The integration of artificial intelligence into judicial systems worldwide has been framed through the lens of performance optimization and efficiency gains. This article argues that such framing is insufficient to capture the nature of AI's impact on justice administration. Drawing on a systematic literature review (n = 89 articles) and integrating sociotechnical systems theory, coproduction theory, and legitimacy frameworks, this study proposes reconceptualizing justice administration as a field of human-AI coproduction. The analysis identified 38 distinct issue categories organized across four dimensions: technology characteristics, the act of judging, law, and society issues. It has been demonstrated that AI goes beyond merely improving existing judicial processes and can significantly reconfigure them, creating new practices, roles, and accountability structures that require analytical frameworks beyond traditional performance metrics. The human-AI coproduction framework provides a theoretical lens that acknowledges AI as an active participant in judicial processes rather than as a passive tool, enabling researchers and practitioners to address emerging challenges, including algorithmic legitimacy, distributed accountability, and the preservation of judicial discretion. The implications for court administrators, legal professionals, technology developers, and policymakers are outlined, and governance mechanisms that balance innovation with the protection of fundamental rights in the judicial system are proposed.