Human Adjudication In The Algorithmic Court: Judicial Independence, Due Process And India's Emerging Framework For Artificial Intelligence In Courts

Full Text Link

Adv. Sonia Saini, Managing Partner, SS Global Law Firm, New Delhi, India

ABSTRACT

Artificial intelligence is no longer confined to court administration. It now structures legal research, summarises records, translates testimony, verifies citations, detects filing defects and, in some jurisdictions, informs assessments of risk. The legal difficulty is not exhausted by asking whether a judge signs the final order. An ostensibly assistive system can shape the evidentiary and doctrinal frame through which a case is understood, while remaining invisible to the parties and difficult to review. This article develops a constitutional account of meaningful human adjudication for Indian courts and evaluates the Supreme Court of India's Draft Regulations for Use of Artificial Intelligence in Courts, 2026 against UNESCO's 2025 Guidelines for the Use of AI Systems in Courts and Tribunals and India's broader AI Governance Guidelines. It adopts a doctrinal and comparative method and advances three findings. First, Articles 14 and 21, the duty to give reasons, open justice, judicial independence and effective appellate review together support a limited right to meaningful human adjudication. Secondly, the Indian draft regulations are substantially aligned with UNESCO and are, in certain respects, more protective: they prohibit risk scoring, solely algorithmic outcomes and opaque systems affecting liberty, and require AI registers, incident reporting and disclosure. Thirdly, material gaps remain. The draft does not create an express right to inspect and contest influential outputs, lacks independent human-rights impact assessment, relies heavily on in-house auditing, leaves 'material assistance' undefined and contains an innovation-oriented presumption that may sit uneasily with the judiciary's rights-protective role. The article proposes a material-influence test and amendments centred on disclosure, contestability, independent evaluation and proportionate remedies. Its central claim is that AI may assist adjudication only where judicial responsibility remains cognitively real, publicly attributable and procedurally challengeable.

Keywords: artificial intelligence; Indian judiciary; judicial independence; procedural due process; human adjudication; UNESCO; algorithmic accountability; open justice.

71. HUMAN ADJUDICATION IN THE ALGORITHMIC COURT- JUDICIAL INDEPENDENCE, DUE PROCESS AND IN

Download HUMAN ADJUDICATION IN THE ALGORITHMIC COURT- JUDICIAL INDEPENDENCE, DUE PROCESS AND IN • 469KB