[Synthetically Drafted | Lawssist-AI]
Key Takeaways
- The Supreme Court set aside judgments from the NCLT and NCLAT dated August 28, 2024, and September 11, 2025, respectively.
- The set-aside orders were based on the tribunals' reliance on non-existent, fake, and AI-hallucinated material presented as legal precedents.
- The Court mandated a "zero-tolerance" policy for citing unverified AI-generated precedents, declaring it professional misconduct for advocates.
- Judicial reliance on fake or hallucinated AI-generated material constitutes a serious lapse, making such decisions void in the eyes of the law.
- The Bar Council of India has been directed to establish guiding principles and disciplinary actions for advocates who present unverified AI-generated material.
Key Background Facts
The case originated from a Section 7 application under the Insolvency and Bankruptcy Code, 2016, filed by Jammu and Kashmir Bank Ltd. against Essel Infraprojects Ltd. (EIL). The appellant, Pooja Ramesh Singh, was a suspended director of EIL, which served as a corporate guarantor for loans availed by Pan India Utilities Distribution Company Ltd. (PIUDCL). Following PIUDCL's defaults, its loan accounts became non-performing assets. The NCLT admitted the bank's application, initiating the Corporate Insolvency Resolution Process against EIL. The NCLAT upheld this decision, dismissing EIL's appeal. During the appeal before the Supreme Court, it was highlighted that the NCLT and NCLAT had relied on numerous fake, non-existent, or AI-generated precedents, some with accurate citations but fabricated paragraphs, to support their findings.Legal Issue Before the Court
The central legal issue before the Supreme Court was whether judgments by lower tribunals, which explicitly relied on non-existent, fake, or AI-hallucinated legal precedents, could be sustained. Specifically, the Court had to determine the validity of decisions where the integrity of the adjudicatory process was compromised by the introduction of unverified AI-generated material. A subsidiary issue concerned the corporate guarantee liability under IBC Section 7, though the Supreme Court deliberately avoided commenting on its merits given the procedural infirmity.Court's Analysis
The Supreme Court undertook an independent examination and confirmed that several judgments cited by the NCLT were either entirely non-existent, wrongly cited, or contained fabricated paragraphs attributed to genuine citations. The Court observed that the appellate tribunal also failed to scrutinize these fake judgments.The Court emphasized that while AI offers extraordinary capabilities and can aid adjudication, its potential to substitute human thinking and decision-making requires extreme caution. It warned that unregulated AI use could "insidiously enter legal practice" and judicial decision-making, highlighting a "human vulnerability to seek comfort in delegation." The judgment stressed that the core of human existence lies in the capacity to think, discern, and arrive at truth through disciplined intellectual exercise and experience, a "Saadhana" that cannot be delegated.
The Supreme Court expressed grave concern over AI's tendency to generate "non-existent, fake, or hallucinated results" when replying to prompts. It analogized the utilization of such fake, AI-generated material as precedents to "the release of methyl isocyanate in the province of law and justice: invisible, insidious, and catastrophic." The Court unequivocally stated that such material contaminates the "very lifeblood of judicial determination."
Therefore, the Court declared a "zero-tolerance mode" for producing, citing, or using AI-generated precedents without verification. It deemed citing such unverified judgments as misconduct for advocates and a serious lapse for judges. Any decision tainted by even an "iota" of fake or hallucinated material, regardless of its bearing on the outcome, violates the sanctity of adjudication and is "no decision in the eyes of the law" and must be set aside. The Court clarified that its ruling does not concern the rightful use of AI but specifically targets the reliance on fake or hallucinated material as valid precedents. It called for "absolute and total control" over AI application, necessitating public policy, enforceable rules, and coordinated efforts between the Bar and the Bench, always maintaining a human in the loop.
Important Observations
The Supreme Court made several critical observations regarding the use of AI in legal practice and adjudication:- "This is yet again a case where the Tribunal relied on non-existent, fake and hallucinated material, generated through Artificial Intelligence (AI), as if it were a precedent in support of its judgment." This highlights a recurring concern about the unchecked use of AI.
- The Court noted that AI's transformative capability, though compelling, demands extra caution to prevent it from replacing human thinking and decision-making in legal processes.
- It philosophically cautioned against the delegation of thinking to AI, asserting that this would have serious consequences for humanity's core capacity to discern truth and justice.
- The judgment analogized the production and utilization of fake AI-generated legal precedents to the "release of methyl isocyanate," emphasizing its unseen yet catastrophic impact on judicial integrity.
- The Court declared that "It is necessary for Courts to adopt a zero-tolerance mode for producing, citing or using AI-generated precedents without verification." This establishes a strict standard for all legal stakeholders.
- It unequivocally stated, "It is a misconduct on the part of an advocate to cite such judgments without verification. Equally, it is a serious lapse if a judge relies on such a fake or hallucinated AI-generated material as precedents in support of the determination." This assigns clear accountability to both the Bar and the Bench.
- Crucially, the Court held that a decision based on fake or hallucinated material "is no decision in the eyes of the law" and must be set aside, even if its influence was minimal, to maintain the "sanctity of adjudication."
- The Court directed the Bar Council of India to establish a committee to deliberate on this issue and formulate guiding principles and disciplinary actions for advocates who misuse AI-generated material.
Outcome
The Supreme Court, for the reasons elaborated, set aside the judgment and orders passed by the NCLT on August 28, 2024, and the NCLAT on September 11, 2025. Consequently, the original Section 7 application, RCP (IB) 6/MB/2023, was restored to its original number before the NCLT. The NCLT was directed to reconsider and dispose of the application expeditiously, preferably within two weeks from the date of the Supreme Court's order, clarifying that no opinion was expressed on the merits of the underlying insolvency dispute. The parties were further instructed to maintain the status quo pending the NCLT's fresh disposal of the application.For the full text of the judgment, refer to 2026 INSC 668.
Practical Implications
This judgment carries profound implications for legal practitioners and the judiciary alike. Advocates are now under an explicit obligation to rigorously verify the authenticity of all legal precedents, particularly when utilizing AI tools for research, as citing unverified AI-generated material constitutes professional misconduct subject to disciplinary action by the Bar Council of India. Judges, in turn, must also exercise heightened scrutiny over cited precedents, as reliance on fake or hallucinated AI inputs can render their decisions void and liable to be set aside. The ruling reinforces the principle of "human in the loop," demanding absolute control and critical assessment over AI applications in adjudication. It signals a new era where technology's benefits must be balanced with strict adherence to truth and accuracy, redefining ethical and professional standards in the context of emerging AI capabilities.Frequently Asked Questions
What constitutes 'AI hallucinations' in legal contexts?
AI hallucinations refer to the generation of non-existent, fake, or fabricated material by Artificial Intelligence systems, often presented as factual or authoritative. In legal contexts, this specifically applies to the creation of non-existent case citations, judgments, or excerpts from judgments that are wrongly attributed to genuine sources. The Supreme Court has likened this phenomenon to a "methyl isocyanate" release in the province of law, emphasizing its insidious and catastrophic potential.
What are the consequences for advocates citing AI-generated fake precedents?
The Supreme Court has unequivocally declared that it is misconduct for an advocate to cite unverified AI-generated fake or hallucinated precedents. The Bar Council of India has been directed to establish guiding principles and disciplinary actions for such violations. This ruling emphasizes the advocate's responsibility to verify the authenticity of all legal material presented in court.
How does reliance on AI-generated fake material affect a judicial decision?
If a judge relies on fake or hallucinated AI-generated material as a precedent, it is considered a "serious lapse," and such a decision is deemed "no decision in the eyes of the law." The Supreme Court mandates that such decisions must be set aside, even if only an "iota" of fake material influenced the decision-making process, as it fundamentally violates the sanctity of adjudication.
What role is the Bar Council of India expected to play regarding AI in legal practice?
The Supreme Court has directed the Bar Council of India to form a committee to deliberate on the issue of advocates submitting fake AI-generated material. This committee is tasked with prescribing guiding principles to prevent such occurrences and establishing disciplinary actions for violations of these norms. The aim is to ensure accountability and maintain professional standards in the era of AI.
Does this judgment prohibit the use of AI in legal research?
No, the judgment explicitly clarifies that it has "no bearing on the rightful use of AI." Instead, it focuses solely on the presentation or reliance on fake or hallucinated material as if it were a court precedent. The Supreme Court acknowledges AI's potential to aid adjudication but stresses the absolute necessity of maintaining human control and verification at every stage.
