The Right to Explanation: Can AI Be a Judge Without Being Accountable?

The integration of artificial intelligence (AI) into judicial processes is frequently justified by the promise of efficiency, consistency and improved access to justice. Yet, when algorithmic systems begin to influence the interpretation of evidence, the application of law or the assessment of legal claims, a fundamental question arises: who is accountable for a decision that cannot be adequately explained? This article examines the constitutional and procedural implications of AI-assisted adjudication in India, with particular reference to Articles 14 and 21 of the Constitution,1 the principles of natural justice and the judicial obligation to provide reasoned decisions. It argues that the central danger of algorithmic adjudication lies not merely in the possibility of technological error, but in the potential separation of decisional power from legal responsibility. Although AI may assist judicial officers in research, translation, document analysis and case management, its outputs must remain subject to meaningful human scrutiny. The article proposes a rights-based framework built around an enforceable right to explanation, disclosure of material AI involvement, independent verification, contestability and identifiable judicial responsibility. It concludes that technological innovation is compatible with constitutional justice only when it strengthens, rather than obscures, the accountability of the institution exercising judicial power.