Landmark Judgment Summary: AI-Generated Fake Case Law and the Sanctity of Judicial Decision-Making

Case: Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. & Anr.

Citation: 2026 INSC 668 | Decision: 2 July 2026

Bench: Justices Pamidighantam Sri Narasimha and Alok Aradhe

The Supreme Court of India delivered a significant ruling on the dangers of relying on artificial intelligence-generated, fictitious judicial precedents. It set aside decisions of the National Company Law Tribunal (NCLT) and the National Company Law Appellate Tribunal (NCLAT), holding that fabricated authorities contaminate adjudication and undermine the rule of law. The judgment establishes that technological efficiency cannot displace the responsibility of judges and advocates to verify the legal foundations of their work.

The dispute arose from insolvency proceedings involving Essel Infraprojects Ltd., which had furnished a corporate guarantee for loans obtained by Pan India Utilities Distribution Company Ltd. from Jammu and Kashmir Bank. Following repayment defaults, the bank initiated proceedings under Section 7 of the Insolvency and Bankruptcy Code, 2016. On 28 August 2024, the NCLT admitted the application, appointed an Interim Resolution Professional and declared a moratorium. The NCLAT upheld that decision on 11 September 2025.

Before the Supreme Court, the appellant challenged the authenticity of six authorities used in the tribunals’ reasoning. Examination revealed three non-existent citations, two genuine citations containing fabricated paragraphs, and another citation wrongly attributed to a different case, again with a non-existent passage. The bank’s affidavit stated that its counsel had not cited these authorities; they had apparently emerged from the NCLT’s own research. The appellate tribunal had also failed to detect the defects.

The central question was whether a judicial determination could remain legally sustainable when its reasoning incorporated fictitious precedents. The Court’s answer was categorical: decisions contaminated by fake or hallucinated legal material must be set aside. Its reasoning did not depend on demonstrating that the fabricated material alone determined the outcome. Even a small amount entering the decision-making process violated the sanctity of adjudication.

This principle gives the ruling particular significance. Ordinarily, appellate scrutiny may distinguish between an immaterial mistake and an error affecting the result. Here, the Court treated fabricated precedent as a fundamental failure of judicial integrity. A court exercises public authority through reasons grounded in actual law. When invented authorities support those reasons, the apparent legal justification becomes unreliable, regardless of how convincing the language appears.

The Court recognised AI’s ability to assist professional work but warned against allowing it to substitute for independent thinking and judgment. It emphasised human control throughout adjudication. Its concern extended beyond inaccurate citations to intellectual dependence: routinely delegating reasoning could weaken the disciplined capacity to distinguish truth from falsehood and reach responsible decisions.

Responsibility attaches to both the Bar and the Bench. The Court characterised an advocate’s citation of unverified AI-generated judgments as misconduct and a judge’s reliance on fabricated material as a serious lapse. It directed the Bar Council of India to constitute a committee to examine the problem and prescribe preventive principles and disciplinary consequences. The ruling therefore connects technological adoption with professional accountability.

The Supreme Court restored the Section 7 application for fresh consideration by the NCLT. It expressly left the substantive insolvency questions undecided, directed expeditious disposal, preferably within two weeks, and required the parties to maintain the existing status quo pending disposal. Accordingly, the judgment invalidated the defective adjudication without deciding whether the underlying debt or guarantee ultimately justified insolvency proceedings.

The broader institutional implications are substantial. Fabricated precedent can deprive litigants of a meaningful opportunity to contest the law supposedly governing their rights. It can also spread through later judgments, multiplying error and increasing litigation costs. These are implications of the ruling: an authoritative judicial format cannot transform machine-generated fiction into law.

The judgment preserves legitimate AI assistance while demanding verified legal foundations. In practical terms, its principle calls for checking the existence of each authority, the accuracy of quoted passages and their relevance to the proposition advanced. Judicial legitimacy depends on authentic sources, independent reasoning and accountable human judgment. The ruling makes those obligations central to the responsible use of AI in courts and tribunals.