New Delhi: In an order highlighting a growing trend of relying on AI-generated non-existent judicial orders, the Supreme Court has set aside orders of both the National Company Law Tribunal (NCLT) and National Company Law Appellate Tribunal (NCLAT) after finding that they quoted “hallucinated” judicial precedents while finalising an insolvency dispute.
A bench of justices P.S. Narasimha and Alok Aradhe called for a zero-tolerance policy by the judiciary toward the use of AI-generated precedents without verification.
For lawyers, the court said, citing such AI-generated decisions “is a misconduct”. Judges, too, would commit a serious lapse if they relied on fake AI-manufactured precedents to support a verdict.
It also held that a judgement based on AI-generated “hallucinated” material is not a “decision in the eyes of the law”.
“Such decisions are to be set aside even if an iota of fake or hallucinated material enters the decision-making process, as it would violate the sanctity of adjudication,” it said, warning of the “invisible, catastrophic” effect it could have if unnoticed.
To address the issue, the court directed the Bar Council of India (BCI) to constitute a committee, deliberate on this issue, frame mandatory regulations and propose disciplinary action for violations.
“The Bar Council must take up this issue with utmost seriousness, deliberate earnestly, and prescribe a guiding principle to prevent such occurrences, along with the disciplinary action that will follow a violation of the norms,” it said.
The Supreme Court has already prepared draft guidelines for the use of AI on both the administrative and judicial sides. The top court has invited suggestions and inputs from relevant stakeholders on the guidelines.
AI’s insidious entry
The court did not completely rule out AI’s assistance in adjudication, but stressed that human control over its application was necessary so that it does not substitute human intelligence, thinking, reasoning and even decision-making.
The present story, however, reveals the contrary, the bench cautioned.
The bench’s concern over AI’s “insidious” entry into legal practice and judicial decision-making was expressed in a 15-page judgement pronounced on Tuesday in an insolvency-related dispute between Jammu and Kashmir Bank Limited and Essel Infraprojects Ltd (EIL).
The judgement was a challenge to NCLAT’s 11 September, 2025 order, affirming NCLT Mumbai’s order to initiate insolvency proceedings against EIL.
During the hearing of the appeal, senior advocate Madhvi Diwan, appearing for the petitioner, informed the court that the tribunals had relied upon six judicial decisions that either did not exist or failed to support the legal arguments advanced by Jammu and Kashmir Bank.
Importantly, both sides submitted that none of the precedents quoted in the judgement were cited by them and that the adjudicating forum, NCLT, obtained them through its own research. Incidentally, the non-existent judgements escaped NCLAT’s scrutiny.
The court’s dependency on technology was addressed in the judgement. So far, courts have seamlessly absorbed technology, making it an integral part of court systems, and have never encountered a problem in its use for the dispensation of justice.
But now, with AI becoming an alternative to the human mind and even decision-making, it was time to be extra cautious, it said.
“Production of fake non-existent material and its utilization as precedents in law is like the release of methyl isocyanate in the province of law and justice: invisible, insidious, and catastrophic by the time anyone notices. It not only contaminates but takes away the very lifeblood of judicial determination,” said the bench.
It then, without “hesitation”, declared that judgements relying on fake judgements produce “hallucinated” results, which upset the integrity of the judicial process.
“It is therefore compelling and necessary to have absolute and total control over the application and usage of AI. The control lies in being two steps ahead of its application and in making deliberate choices about when and where to apply,” the bench advised.
The judgement weighed in on the serious consequences of the delegation of human thinking to technology, which it felt, if lost, would result in a “loss of everything”.
“Wisdom and foresight compel us to recognise human vulnerability to seek comfort in delegation, but if thinking is delegated and it forms a habit, it will have serious consequences for the core of human existence, which lies in its capacity to think – to discern the distinction between what is right and what is wrong, truth and falsehood, virtue and vice, dharma and adharma,” the court said.
This capability is built and strengthened over a period of time with intellectual exercise, coupled with experience, and is not superimposed by birth, it said. It enables humans to “choose between competing values, as well as to take hard decisions with courage and conviction, and to bring about a beautiful balance between the need for order and the quest for justice.”
This struggle to arrive at truth is a saadhana, without which we lose “the capability to discriminate between what is right and what is wrong”.
(Edited by Sugita Katyal)
Also Read: AI in courts: How India’s draft rules stack up against the EU, US and China