SC flags usage of AI-generated fake judgements as precedents in law, likens it to 'release of…
The Supreme Court of India has issued a zero-tolerance policy against the use of AI-generated fake legal precedents following an NCLT Mumbai insolvency case
- Incident date
- Jul 2026
- Target
- NCLT Mumbai
In July 2026, the Supreme Court of India took decisive action against the submission of AI-generated, hallucinated legal precedents following an insolvency dispute involving NCLT Mumbai. The court declared that any decision influenced by such fabricated material is invalid, emphasizing the necessity of maintaining integrity in judicial processes.
What happened
The issue surfaced during an appeal regarding an insolvency application involving Pooja Ramesh Singh, Jammu and Kashmir Bank Ltd, and Essel Infraprojects Ltd. Upon review, the Supreme Court discovered that the NCLT Mumbai order relied on several non-existent legal precedents to justify its ruling. These fabrications included entirely invented case names and paragraphs that were falsely attributed to genuine citations. Specifically, the NCLT cited cases such as ICICI Bank Ltd vs Urban Infrastructure Real Estate Ltd (2019) and Sarbjit Singh vs Union Bank of India (2022), both of which were found to be completely non-existent.
While the respondent, Jammu and Kashmir Bank, clarified that their counsel had not provided these citations, noting that the NCLT had retrieved them through its own research, the Supreme Court ruled that the origin of the error did not excuse the violation of the rule of law. The bench likened the introduction of fake material into the judicial system to the release of methyl isocyanate, underscoring the danger of polluting the legal record.
The Supreme Court has since mandated a zero-tolerance policy for both the Bar and the Bench regarding the citation of AI-generated material. The Bar Council of India has been directed to form a committee to establish guiding principles and disciplinary actions to prevent future occurrences. While the court explicitly stated that this ruling does not discourage the rightful use of AI in legal research, it underscored that human oversight must remain at every stage of the adjudication process to ensure that only verified legal authority informs court decisions.