Author: Devalay Dey
College: Maharishi University of Information and Technology, Noida
To the Point
Artificial Intelligence (AI) is rapidly becoming an important tool in legal research, drafting and information management. However, its use in judicial proceedings creates serious concerns when AI-generated information is treated as genuine legal authority without verification.
The Supreme Court of India addressed this issue in PoojaRamesh Singh v. Jammu & Kashmir Bank Ltd. & Anr., 2026 INSC 668, decided on 2 July 2026. The Court set aside orders of the National Company Law Tribunal (NCLT) and the National Company Law Appellate Tribunal (NCLAT) after finding that the NCLT had relied upon six citations that were either non-existent or did not support the propositions attributed to them. The material appeared to have been generated through AI.
The Supreme Court made it clear that AI may assist the legal system, but it cannot replace human verification, judicial reasoning or responsibility. The judgment therefore establishes an important principle: technology may assist adjudication, but legal authority must always remain subject to human verification and judicial control.
Use of Legal Jargon
The controversy involves several important legal concepts:
Judicial Precedent: A previous judicial decision used as an authoritative source for determining a legal issue.
AI Hallucination: A situation where an AI system produces information that appears credible but is factually incorrect, fabricated or non-existent.
Ratio Decidendi: The legal principle necessary for deciding a case and capable of being relied upon in subsequent cases.
Adjudication: The formal process through which a court or tribunal determines legal rights and liabilities.
Rule of Law: The fundamental principle that governmental and judicial authority must operate according to established law rather than arbitrary decision-making.
Judicial Integrity: The requirement that judicial decisions be based upon authentic evidence, law and legitimate legal reasoning.
Professional Misconduct: Conduct by a legal practitioner that violates professional standards or duties imposed upon advocates.
Natural Justice: Fundamental procedural principles requiring fairness, impartiality and a reasonable opportunity to be heard.
In the present context, the most important principle is that a precedent must actually exist and must support the proposition for which it is cited. An AI-generated citation cannot acquire legal authority merely because it appears convincing.
The Proof
The Supreme Court judgment provides the strongest evidence for the legal position.
The dispute originated from insolvency proceedings involving Essel Infraprojects Ltd. The NCLT had relied upon six judicial citations while dealing with the insolvency proceedings. Subsequent examination revealed that three cited judgments did not exist, while the remaining citations either did not contain the propositions attributed to them or were incorrectly represented.
The matter reached the Supreme Court in Civil Appeal No. 11950 of 2025. A Bench comprising Justice Pamidighantam Sri Narasimha and Justice Alok Aradhe examined whether a decision based upon such fabricated legal material could legally survive.
The Supreme Court answered in the negative. It held that where fake or hallucinated material enters the decision-making process, the adjudicatory process itself becomes compromised. The Court consequently set aside both the NCLT and NCLAT decisions and restored the matter to the NCLT for fresh consideration.
Importantly, the Court did not prohibit the legitimate use of AI. Instead, it emphasised that AI should operate under meaningful human supervision. The judgment recognises AI’s potential to assist with both routine and intellectual legal work while insisting that ultimate control over adjudication must remain with human decision-makers.
The Court also directed the Bar Council of India to constitute a committee to examine the issue of advocates submitting fake or hallucinated AI-generated legal material and to consider appropriate guiding principles and disciplinary consequences.
This establishes a significant distinction:
AI-assisted legal research is not inherently unlawful; relying upon unverified AI-generated material as genuine law is unacceptable.
Abstract
The increasing integration of Artificial Intelligence into legal practice presents both opportunities and risks. AI can assist advocates, judges, researchers and tribunals by processing large volumes of information and facilitating legal research. Nevertheless, generative AI systems may produce fabricated authorities, incorrect quotations or imaginary judgments while presenting them in an apparently authoritative manner.
The Supreme Court’s decision in Pooja Ramesh Singh v. Jammu & Kashmir Bank Ltd. represents an important judicial response to this emerging problem. The Court held that an adjudicatory decision founded upon fake or hallucinated legal material cannot be sustained because it undermines the integrity of the judicial process and the rule of law.
The judgment does not reject technological innovation. Rather, it establishes a human-centred approach in which AI may function as an assisting mechanism while responsibility for legal verification and adjudication remains with human professionals.
The decision is particularly significant for Indian legal education and professional practice because it reinforces the traditional duty of lawyers to verify authorities before placing them before a court. It also demonstrates that technological advancement cannot dilute established standards of professional responsibility.
Case Laws
1. Pooja Ramesh Singh v. Jammu & Kashmir Bank Ltd. & Anr., 2026 INSC 668
This is the principal case concerning AI-generated fake precedents. The Supreme Court found that the NCLT had relied upon six problematic citations, several of which were non-existent or incorrectly attributed. The NCLAT subsequently affirmed the decision.
The Supreme Court set aside both orders and directed fresh consideration of the insolvency proceedings. It declared that even limited reliance upon fake or hallucinated material can compromise the sanctity of adjudication. The Court simultaneously clarified that its criticism was directed at fabricated legal material being presented as precedent, rather than the legitimate use of AI as a technological tool.
2. State Bank of India v. Shree Ram Urban Infrastructure Ltd., (2020) 16 SCC 728
This decision was among the authorities considered in the proceedings. It demonstrates the importance of relying upon genuine judicial precedents while deciding questions arising under insolvency law. The case is particularly relevant because the present dispute also arose within the framework of the Insolvency and Bankruptcy Code, 2016.
3. ICICI Bank Ltd. v. Sree Urban Infrastructure Real Estate Ltd., (2019) 16 SCC 528
This decision was also cited among the authorities connected with the legal dispute. Its relevance demonstrates the importance of accurate citation and contextual application of precedent in insolvency adjudication.
4. Everest Kento Cylinders Ltd. v. Union of India, (2015) 2 SCC 1
This Supreme Court decision was another authority identified in the case record. Its inclusion further illustrates that legal precedents must be authentic and correctly applied rather than merely appearing authoritative through an AI-generated response.
Conclusion
The Supreme Court’s judgment in Pooja Ramesh Singh v. Jammu & Kashmir Bank Ltd. marks an important development in Indian legal jurisprudence concerning Artificial Intelligence.
The judgment does not represent a rejection of AI. Instead, it establishes a principle of responsible technological integration. AI can assist lawyers and judges in research, document analysis and other tasks, but it cannot become a substitute for legal judgment, verification and accountability.
The case also reinforces the professional responsibility of advocates. Before citing any authority, a lawyer must verify that the judgment exists, confirm its citation and examine whether the judgment actually supports the proposition for which it is being relied upon.
The larger lesson is clear: AI may accelerate legal research, but it cannot replace legal responsibility.
For the judiciary, the challenge will therefore be to develop a framework in which technological innovation and judicial independence coexist. The Supreme Court’s insistence upon a human being “in the loop” provides an important foundation for this future.
Ultimately, the legitimacy of the judicial process depends not upon how advanced the technology is, but upon whether the law relied upon is genuine, verifiable and correctly applied.
FAQ
1. Does the Supreme Court prohibit lawyers from using AI?
No. The judgment does not prohibit legitimate AI use. The Court recognised that AI can assist legal work but stressed that its output must remain subject to human verification and control.
2. What is an AI hallucination in legal research?
An AI hallucination occurs when an AI system generates information that appears authentic but is actually false. In legal research, this may include imaginary cases, incorrect citations or fabricated quotations.
3. What happens if a court relies upon an AI-generated fake precedent?
According to the Supreme Court, a decision affected by fake or hallucinated legal material cannot be sustained and may have to be set aside because the integrity of adjudication has been compromised.
4. Who is responsible for verifying AI-generated legal research?
The ultimate responsibility remains with the human legal professional or decision-maker using the material. AI output should therefore be independently checked against authentic judgments and statutory sources.
5. Why is this judgment important for law students?
The judgment highlights that legal research is not simply about obtaining information. A law student must verify authorities, understand their ratio decidendi, examine the factual context and ensure that the precedent genuinely supports the legal proposition being advanced.
6. Can AI replace judges or lawyers?
The judgment strongly supports a human-controlled approach. AI may assist legal professionals, but adjudication requires human reasoning, responsibility and judicial discretion. The Court specifically emphasized maintaining human control over adjudication.
