The Supreme Court's ruling on AI-hallucinated legal precedents raises fundamental questions about the role of artificial intelligence in India's justice delivery system. Critically examine the implications for judicial integrity and advocate accountability.

Q. The Supreme Court's ruling on AI-hallucinated legal precedents raises fundamental questions about the role of artificial intelligence in India's justice delivery system. Critically examine the implications for judicial integrity and advocate accountability. (15 marks, 250-350 words)

In Pooja Ramesh Singh v. J&K Bank (2026) [1], the Supreme Court set aside NCLT and NCLAT orders in the Essel Infraprojects insolvency for resting on fabricated, AI-generated citations. The ruling is a timely corrective, though its enforcement machinery remains untested.

AI's legitimate role in justice delivery - The Court accepted AI as an assistive tool for legal research and drafting, valuable amid heavy tribunal pendency, but insisted "human control over adjudication must remain total and absolute" [1]. - The standard is thus regulated adoption, not prohibition — a workable norm for e-courts and tribunal modernisation.

Implications for judicial integrity - A decision resting even partly on hallucinated material is "no decision in the eyes of the law" [1] — offending natural justice and the rule of law under Article 14. - Stare decisis presupposes authentic, verifiable rulings; fake precedent silently corrupts the chain of authority — the Court's "methyl isocyanate of law" metaphor. - Zero tolerance binds Bench as well as Bar. Quasi-judicial bodies like NCLT, constituted under Section 408, Companies Act 2013 [2] and lacking High Court–grade research support, are structurally more exposed to such shortcuts.

Implications for advocate accountability - Citing unverified AI output was held to be professional misconduct, placing the verification duty squarely on counsel as officers of the court. - The Bar Council of India — statutory regulator exercising disciplinary jurisdiction and prescribing professional conduct standards under the Advocates Act, 1961 [3] — was directed to constitute a committee and frame guidelines.

Critical limitations - India has no dedicated AI statute; a BCI committee is self-regulation, weak on deterrence. - Liability of AI vendors and of non-verifying adjudicators stays undefined. - Remand for fresh hearing adds delay, straining IBC's time-bound resolution.

The judgment rightly locates the remedy in human responsibility rather than technological retreat. A durable framework needs mandatory citation-verification protocols, AI-literacy training for judges and members, and audited legal-research databases. Anchored in institutional integrity, technology can then widen access to justice without diluting it.

(~320 words)

Sources: 1. Pooja Ramesh Singh v. Jammu & Kashmir Bank Ltd., 2026 INSC 668 (SC, 2 July 2026) — setting aside of NCLT/NCLAT orders, "no decision in the eyes of the law", zero tolerance, human control over adjudication, professional misconduct, BCI direction 2. About NCLT, National Company Law Tribunal — NCLT constituted under Section 408, Companies Act 2013; quasi-judicial adjudicating authority under IBC, 2016 3. About the Bar Council of India — statutory body under the Advocates Act, 1961; prescribes professional conduct standards and exercises disciplinary jurisdiction