On 2 July 2026, the Supreme Court of India delivered judgment in Pooja Ramesh Singh v. Jammu and Kashmir Bank Limited and Anr., Civil Appeal No. 11950 of 2025, reported as 2026 INSC 668. The Court held that a judicial or quasi-judicial decision relying on fabricated case citations generated by artificial intelligence is unsustainable in law and must be set aside. The ruling is final and binding on all courts and tribunals in India under the doctrine of precedent.
The Supreme Court grounded the ruling in established principles of natural justice and valid adjudication: a decision premised on legal authority that does not exist in law has no proper foundation and cannot stand. The Court applied no AI-specific statute; it extended existing doctrine that an order vitiated by a fundamental error in its stated reasoning must be set aside. By applying that doctrine to AI hallucinations, the Court established that reliance on non-existent citations is a cognizable defect warranting setting aside, not merely a procedural irregularity curable on remand.
Legal practitioners appearing before Indian courts, tribunals, and regulatory bodies must verify every cited decision before including it in a written submission or placing it on record. Practitioners using AI tools for Indian legal research carry an independent professional duty to confirm that each cited decision exists, is accurately described, and remains good law. Banks, non-banking financial companies, fintech operators, and other regulated entities involved in administrative or commercial proceedings in India should audit their in-house and external legal teams for AI citation practices and require verification protocols before any AI-assisted research is submitted.
The ruling leaves open the question of remedies where AI-generated material has already been placed on record and informed a now-final order. It also raises practical questions for the National Company Law Tribunal, Securities Appellate Tribunal, and other specialized bodies that have experimented with AI-assisted research tools. Litigants in affected proceedings may seek review on the authority of this judgment.
Licentium monitors judicial and regulatory developments affecting AI governance in emerging markets and advises clients on AI-related legal risk. Work we undertake includes AI legal risk assessments for financial institutions and technology companies, cross-border litigation support, regulatory compliance advice for fintech operators in India, AI governance program design, and jurisdictional briefings on AI adjudication risk.