Emerging Legal Risks in India
Author: Satyam Dubey
College: Symbiosis Law School Nagpur
Abstract
Generative Artificial Intelligence (AI) has revolutionized legal research, drafting, and legal practice. Lawyers and litigants can leverage AI tools to search for authorities, summarize judgments, draft pleadings and analyze complex legal issues. But the same technology poses a big and growing danger: the creation of ghostly or bogus case laws, judicial decisions, citations, quotations, and legal propositions. This “fake case law” may be included in court filings and other legal documents, and may mislead judges, parties and attorneys.
This article explores how the Indian legal system is facing new challenges with the rise of fabricated case law created by AI. It examines the consequence of legal hallucinations on judicial decision making; on professional ethics; on access to justice; on procedural justice; and on the integrity of the legal profession. The article also examines the current legal and institutional landscape in India, such as the responsibilities of advocates, judicial accountability for relying on authorities generated by AI, principles of natural justice, and the evidentiary and procedural implications of using unreliable AIgenerated material. Prior Indian judicial rulings on ethical issues in legal practice, misleading courts, suppression or misrepresentation of material facts and the reliability of technology are explored to gauge their applicability in the context of AI-integrated legal research.
The article concludes that AI should not be a legal source of information in its own right, but should be used as an assistive technological tool. It ends with recommendations for enhanced verification systems, professional guidelines, judicial awareness, AI literacy, disclosure and accountability systems, to help prevent technological innovation from compromising the trust and functioning in the Indian justice system.
To the Point
A.I. is increasingly being utilized in legal research and legal drafting. These generative AI systems can generate seemingly convincing legal arguments, citations, case names, judicial quotations, and statutory references in seconds. The biggest challenge is that an AI system can create information that may seem legal but isn’t. An AI-generated fabricated judgment might include an invented case number, a fake citation, a made-up paragraph number, a nonexistent judicial remark or a quotation that was never actually uttered by a court. Traditionally, law writing is a skill that is highly dependent on precedent, and the production of such fiction has significant repercussions in the judicial process. The danger is especially great, given that the law is not just a source of information. Precedents impact on the construction and application of laws. If accepted without verification, the precedent could lead to incorrect legal reasoning and impact the procedure’s results.
The issue is not just technological inaccuracy, but also. It presents fundamental issues regarding the responsibility of the profession, judicial independence, fairness in procedure, access to justice and public trust in the administration of justice. Don’t automatically ban AI usage in law – that’s not the answer! Instead, a system should be in place for courts and legal practitioners to verify AI products and ensure that they are used properly in legal cases.
Use of Legal Jargon
The creation of ‘fake law’ by AI also links with several fundamental principles of Indian legal jurisprudence. The principle of stare decisis, professional ethics, duty of candour, natural justice, procedural fairness and administration of justice comes into play. The legal profession has a responsibility on advocates to help the court to administer justice. As a general rule, the products of AI should not be considered a source of legal information by itself, even if it seems correct or uses a
promising-looking citation. Legal research involves checking against primary sources of the law, such as decisions of competent courts, laws and recognized legal databases. The principle of stare decisis presupposes the existence of a true judicial decision which can be used to establish precedent. A fabricated judgment cannot be used as a precedent since it is not a decision of a lawfully constituted court. Therefore, there can be no binding or persuasive precedent based upon no authority whatsoever.
The idea of natural justice also crops up where an artificial authority impedes the effective response of an opposing party to a legal argument. Where one party’s reliance on a non-existent judgment, and the court assumes the authority is real, the opponent could be at an unfair procedural disadvantage.AI hallucinations could also give rise to further questions under the Advocates Act, 1961 and the professional standards of advocates. When an advocate deliberately gives a court false information, the actions can be professionally and disciplinary significant. But the exact legal effect will depend on the knowledge, intent, circumstances and applicable rules of procedure and professional conduct of the advocate.
Understanding the difference between AI-aided legal research and AI-driven legal reasoning, therefore, is crucial. While AI can help a legal professional find or organize information, the final responsibility for the correctness of each submission is with the human legal professional.
The Proof
The danger of faked legal grounds isn’t simply hypothetical. Predictive and generative AI systems are created to generate likely language – not to make the determination that each statement reflects an authentic legal source. Thus, an AI system may generate a citation that looks like a real Indian case, but contains incorrect case names, dates, quotations, courts, or paragraph numbers. The legal relevance of this problem is very important since judicial proceedings rely on the authenticity of authorities presented to the court. A fabricated judgment differs from a regular factual error in that it can wrongly ascribe such constitutional court legal reasoning to the court, giving the impression of a legal proposition when there is none. The seriousness of the problem is increased if fabricated authorities are added to pleadings without verification. If it is entered as part of the judicial record, then opposing counsel or the judges may have to expend extra time to establish its authenticity.
This can make litigation more expensive and slower down the process. The problem also presents a general institutional risk. The integrity of submissions by members of the Bar is the basis upon which Courts rely. The continued use of un-sourced AI-generated content may undermine confidence in the Bench and the Bar and have an impact on public confidence in the administration of justice.
However, the presence of AI-related harms doesn’t necessarily mean that all uses of generative AI are illegal. AI can be helpful in research, summarization, drafting, translation and organization of legal text.AI can be helpful for research, summarization, drafting, translation and organization of legal text. The key to protecting everyone is for each authority to be independently verified before it’s used in court.
Case Laws
1.Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. & Anr. (2026)
The Supreme Court set aside the NCLT and NCLAT decisions after finding that the NCLT had relied upon fake, non-existent, and AI-generated case law. The Court adopted a zero-tolerance approachtowards citing or relying upon AI-generated precedents without verification and held that such reliance undermines the integrity of judicial decision-making. The Court also directed the Bar Council of India to formulate guidelines and disciplinary measures regarding the use of fake AI-generated legal authorities.
2.The Supreme Court in State of Uttar Pradesh v. Raj Narain (1975) has emphasized on the significance of the rule of truth and fairness in the judicial proceedings that must also be supplied to the court with reliable material. For AI-generated legal writing, this judgment underscores the need for precision and clarity in the information it offers to courts.
3.Dhanraj Giri Hospital v. Haribhai Lakhani (2017), the Supreme Court laid emphasis on the role of the advocates in helping the court in the process of administration of justice. The case is significant to the professional responsibility of lawyers to ensure submissions to courts are responsible and accurate.
4.In O.P. Sharma v. High Court of Punjab & Haryana (2011), the Supreme Court outlined the principles and values that should be upheld by lawyers, and reiterated that lawyers have duties not just to their clients. This principle is especially pertinent in the context of technology tools that can produce unreliable legal authorities, used by lawyers.
5.Himalayan Coop. In Group Housing Society v. Balwan Singh (2015) the Supreme Court
emphasized on the role of the lawyer and the need for purity in the legal profession. It offers a helpful template for thinking about the application of current professional duties with respect to new technologies, including generative AI, in the future.
6.In the case of State of Maharashtra v. Dr. Praful B. Desai, 2003, the Supreme Court has acknowledged the importance of technology in the administration of justice, especially in regard to the technological method of recording evidence. The case demonstrates that Indian courts do not have an animosity towards technological innovations. But this embrace of technology should be consistent with the demands of reliability, fairness, and justice.
7.Anvar P.V. v. P.K. Basheer (2014) — The Supreme Court of India considered the admissibility of electronic records as evidence and highlighted the need to abide by statutory conditions for the admission of electronic material. The case itself does not focus on generative AI, but it does highlight the importance of authenticity and legal reliability as it relates to wider issues of treating technologically-generated material in legal proceedings.
Conclusion
AI-generated fake case law is a new technological hurdle to an age-old and core principle of the justice system: decisions should be made based on legitimate laws and credible legal sources.
GenAI can have a significant impact on elevating legal research and enhancing legal efficiency. But it can generate convincing but inaccurate legal authorities, which can be problematic if users don’t double check its work. A fabricated judgment can seem convincing, but never have been recorded, which could lead to deceptive courts, burdening judicial time, rising litigation expenses, and loss of confidence in the legal profession. The principles of Indian law have already a number of elements that can be applied to solve part of this problem. Professional obligations to the court; fairness and candour in judicial proceedings; and the importance of judicial precedent.
These principles can offer a framework for balancing the regulation of irresponsible AI use with the need to not impede legitimate technological innovation. The response should cover the requirement to verify AI-generated authorities, professional advice for advocates, AI-literacy training programs for lawyers and judges, institutional verification processes, responsible-use policies, and suitable consequences for knowingly providing false authorities.
Therefore, the principle should be embraced by courts and legal professionals that while AI can aid in legal reasoning, it does not eliminate legal accountability. All AI-generated case citations, quotations, statutory provisions and judicial propositions shall be independently verified against the primary source document before being presented to court. The question is not whether the Indian judicial system should rely on AI, but how it can rely on AI, without compromising the values of accuracy, accountability, professional integrity, judicial independence and the rule of law.
FAQs
Q1. AI-generated fake case law?
AI-generated case law that is fake or non-existent, case citations, quotations, paragraph references, legal authorities generated by systems of generative AI.
Q2. Why is it that AI can make fake judgments?
Generative AI is capable of predicting and creating text from patterns it was trained on and was instructed to do so. Does not necessarily check all citations against a true judicial database for an answer.
Q3. Relying on an AI-generated case citation?
An advocate should not rely on an AI-generated case citation without using a reliable primary or authoritative legal source to confirm the existence, authenticity, citation and legal proposition of the judgment.
Q4. If it is a case that is fabricated, is it a precedent?
No. The absence of a judicial decision means it could not be a precedent since it never came to a competent court.
Q5. Will AI-made fake case laws be held to account by lawyers?
Absolutely. Using AI doesn’t remove the professional responsibility from the advocate’s shoulders. Consequences would be dependent on the situation, such as if the advocate had been aware or reasonably should have been aware that the authority was false and what the professional and procedural rules would be.
Q6. Should the practice of law be prohibited of AI usage?
A blanket ban could unnecessarily limit positive uses of AI. Responsible AI with human verification, professional standards, transparency and accountability is a better way.
Q7. How should AI be used to conduct legal research?
AI should be used to help with legal research and drafting, not as a source of information. All case citations, quotations, statutes and legal propositions must be checked separately from an authentic source before they are relied upon.
References
1. Constitution of India, 1950.
2. Advocates Act, 1961.
3. Bharatiya Nagarik Suraksha Sanhita, 2023.
4. Bharatiya Sakshya Adhiniyam, 2023.
5. O.P. Sharma v. High Court of Punjab & Haryana, (2011) 6 SCC 86.
6. Dhanrajgiri Hospital v. Haribhai Lakhani, (2017) 2 SCC 744.
7. Himalayan Coop. Group Housing Society v. Balwan Singh, (2015) 7 SCC 373.
8. State of Maharashtra v. Dr. Praful B. Desai, (2003) 4 SCC 601.
9. Anvar P.V. v. P.K. Basheer, (2014) 10 SCC 473.
10. Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal, (2020) 7 SCC 1.
