Back
AI Disclosure Rules Will Burden Lawyers, Justice Prathiba Singh Warns
In a candid assessment at the International ADR Conference,
, organised by
, Justice Prathiba M Singh of the
cautioned that imposing mandatory artificial intelligence (AI) disclosure norms on lawyers would add yet another layer of complexity to an already technology-heavy litigation process. Speaking on the theme
"ADR Pathways to Legal Harmony,"
the judge acknowledged the growing role of AI in legal practice but suggested that the adversarial system itself offers sufficient safeguards against abuse, making rigid disclosure requirements premature.
Her remarks come at a time when courts worldwide are grappling with how to regulate the use of generative AI tools—ranging from ChatGPT to specialised legal research platforms—that can draft pleadings, summarise judgments, and even predict case outcomes. India, with its rapidly digitising judiciary and ambitious e-courts project, is no exception. The already mandates electronic filing and OCR (optical character recognition) of documents, reflecting a broader push toward paperless litigation. Yet Justice Singh questioned whether adding disclosure obligations on top of these existing burdens would genuinely enhance integrity or merely create bureaucratic templates.
Background: The AI Disclosure Debate
The debate over AI disclosure in legal proceedings has intensified as lawyers increasingly rely on generative tools to prepare cases. Several jurisdictions, including the United States, have seen high-profile incidents where attorneys submitted filings containing fictional case citations generated by AI, prompting judges to adopt standing orders requiring disclosure of AI use. In Canada and the United Kingdom, courts have issued practice directions mandating lawyers to certify whether AI was used in drafting documents and to verify the accuracy of the output.
India, however, has yet to adopt a formal framework. The has issued cautionary remarks but stopped short of imposing blanket disclosure rules. In this vacuum, individual high courts—most notably the —have taken an approach, with some judges asking lawyers to confirm that AI-generated content has been verified. Justice Singh’s comments highlight the tension between the need for accountability and the practical realities of legal practice, where turnaround times are short and cost pressures intense.
Key Developments: A Judge’s Pragmatic View
Responding to a question from Deepak, a conference participant, about whether disclosure norms were necessary until an institutional framework was introduced, Justice Singh was blunt:
"I think disclosure norms is just going to make lawyers’ life more difficult."
She elaborated on the technological requirements that lawyers already navigate under the and the electronic filing system, pointing to the mandatory OCR and e-filing steps that consume valuable time.
"They have to OCR documents, e-file documents and at the end of the day, even if you ask for disclosures, they all become
, right?"
she said. In her view, requiring a standardised declaration would likely produce boilerplate language that offers little genuine protection, while adding another compliance checkbox for practitioners.
More significantly, Justice Singh argued that India’s inherently guards against misleading AI-generated content.
"If one party relied on such material, the opposing party was likely to flag it before the court,"
she explained. Because both sides are incentivised to scrutinise each other’s submissions, errors or fabrications are likely to be exposed in the ordinary course of litigation. This self-correcting mechanism, she suggested, reduces the need for a proactive, court-imposed mandate.
Legal Analysis: Balancing Innovation and Accountability
Justice Singh’s observations raise a critical question: is the adversarial process a sufficient safeguard against AI misuse, or does it create a reactive system that only catches problems after they cause harm? While it is true that opposing counsel will often identify inaccuracies, there are scenarios where both parties may rely on the same flawed AI output—for instance, when a tool generates a plausible but incorrect interpretation of a statute that neither side detects during drafting.
Moreover, the risk of “” is real. If disclosure becomes a mere formality, it loses its intended effect of prompting lawyers to verify AI-generated content and take responsibility for its accuracy. The judge’s scepticism, however, is not unfounded. In many courts, similar requirements—such as signing a —have devolved into stamp-and-submit exercises. The real challenge lies in designing a framework that is substantive rather than symbolic.
Another layer is the evolving regulatory landscape. The recent recommendation of the to consider a rule requiring lawyers to disclose AI use in pleadings, without specifying consequences, suggests that a formal proposal is in the works. Justice Singh’s remarks foreshadow the debate that will likely accompany any such rule: how to enforce it, what level of detail to demand, and how to balance transparency against the practical burden on practitioners, especially those in smaller firms or with limited technological infrastructure.
Impact on Legal Practice
For the legal community, Justice Singh’s words are a reminder that the adoption of AI in litigation is still in its infancy. The "churning" period she refers to—when tools are being tested, errors are being made, and best practices are yet to crystallise—demands patience rather than premature regulation. Her call to
"let it churn a little more"
suggests that judges themselves are still becoming comfortable with the technology and that guidelines should emerge organically from experience.
In the interim, lawyers should consider adopting voluntary best practices, such as verifying AI-generated citations and clearly stating the extent of AI assistance in their filings, if only to avoid judicial displeasure. Courts, for their part, may find it useful to issue guidance that distinguishes between using AI for research and using it for substantive drafting, and to encourage transparency through practice directions that are pragmatic rather than punitive.
The ’s existing technological mandates—like OCR and e-filing—already impose a baseline of digital competence. Adding AI disclosure on top requires careful thought about who bears the burden of proof and where the line between assistance and authorship lies. Justice Singh’s observation that such norms “just make lawyers’ life more difficult” resonates with practitioners who are already juggling multiple compliance demands in a fast-paced environment.
Conclusion
Justice Prathiba Singh’s remarks inject a pragmatic voice into the increasingly heated debate over AI regulation in legal practice. By emphasising the adversarial system’s inherent checks and the risk of formulaic disclosures, she counsels against a hasty regulatory rush. Her call for guidelines to be developed and published after a period of observation reflects a measured approach that many in the legal profession will welcome.
As courts and regulators continue to ponder the future of AI in litigation, the balance between innovation and accountability remains delicate. While disclosure norms may eventually become inevitable, Justice Singh’s warning underscores the need for thoughtful design—one that avoids burdening lawyers with pointless paperwork while still ensuring that the integrity of legal proceedings is upheld. For now, the churn continues, and the legal community watches with interest as India charts its course in the age of artificial intelligence.
