Role of Artificial Intelligence in the Criminal Justice System: A Critical Legal Study of Investigation and Judicial Processes

Abstract

Artificial Intelligence (AI) has moved from the margins of legal administration to the very centre of contemporary criminal justice reform. From predictive-policing platforms and facial-recognition surveillance networks used in criminal investigation, to risk-assessment algorithms such as the Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) deployed at the sentencing stage, and to court-assistance tools such as India's Supreme Court Portal for Assistance in Court's Efficiency (SUPACE) and the Supreme Court Vidhik Anuvaad Software (SUVAS), AI now performs functions that were, until recently, the exclusive preserve of human legal actors (Berk, 2021; Kaur & Kaur, 2024). This article undertakes a critical legal study of the role of AI across the two principal stages of the criminal justice process — investigation and adjudication — situating the analysis within constitutional guarantees of due process, equality, privacy and protection against self-incrimination (Articles 14, 20(3) and 21 of the Constitution of India), and within comparable due-process jurisprudence abroad, most notably State v. Loomis, 881 N.W.2d 749 (Wis. 2016) and Justice K.S. Puttaswamy (Retd.) v. Union of India, (2017) 10 SCC 1. Employing a doctrinal, analytical and comparative methodology, the study examines statutory and constitutional provisions, judicial pronouncements, and institutional policies governing the use of AI in evidence collection, forensic analysis, bail and sentencing prediction, and judicial case management. The analysis is organised around three research objectives concerning investigation, judicial decision-making, and the legal-ethical implications of algorithmic governance. The study finds that while AI offers demonstrable gains in efficiency, case-clearance rates and access to justice, its unregulated deployment carries a serious risk of eroding due process, entrenching discriminatory bias, and diluting the accountability structures on which criminal justice legitimacy depends. The article concludes by recommending a graduated, rights-based regulatory framework for AI in the Indian criminal justice system, drawing on the risk-based model of the European Union's Artificial Intelligence Act and on emerging Indian data-protection jurisprudence.