Open Access Research Article
VERIFIED, NOT MERELY CITED: REFORMING ACCOUNTABILITY FOR GENERATIVE AI USE BY ADVOCATES AND JUDGES IN INDIA
Abstract
A fabricated judgment looks exactly like a real one: it
carries a case name, a neutral citation and a confident ratio,
yet it exists nowhere. As generative AI enters legal
practice, Indian courts have
begun to treat the filing
of such authority as professional misconduct, most notably in Pooja Ramesh Singh v. Jammu and
Kashmir Bank Ltd. (2026 INSC 668), and have set aside orders that rested on
non-existent precedent. Yet India still has no framework that says who is
answerable, for what state of mind, and with what consequence. This paper
argues that the present position is incomplete in four respects: the disciplinary law for advocates
is generic and blunt; the accountability route for judges is
fragmented and ill-suited to verification failures;
the question of disclosure has been left to individual courts; and the remedy
of treating a tainted decision as “no decision in the eyes of the law” sits uneasily
with finality and res judicata. Its answer is a single organising
principle, a non-delegable duty of verification, applied in a role-sensitive
manner to advocates and judges alike, with consequences scaled to culpability
rather than to outcome. For advocates, the paper proposes
a technology-neutral amendment to the Bar Council of India Rules under Section
35 of the Advocates Act, 1961, a four-tier culpability ladder and a referral
protocol from courts
to Bar Councils. For judges,
it proposes a graded
ladder that begins with correction and training. It recommends a verification
certificate in place of a blanket AI-use disclosure rule, assesses the Kerala
High Court’s 2025 policy for the district
judiciary, and offers
a point-of-reliance rule and a national model policy for subordinate courts. Finally, it
proposes a materiality test, with remand as the preferred remedy, so that courts
nullify where contamination is real and correct where it is not. The method is doctrinal and comparative, drawing
on recent Indian
decisions and on practice
in the United States, the United Kingdom, Canada and New Zealand.
Published Paper
Author Details
Authors: RASHAP BANSALRegistration ID: 1013396 | Published Paper ID: IJLRA13396Approved ISSN: 2582-6433 | Country: Delhi, IndiaPage No.: 1594-1618
Article Information
VERIFIED, NOT MERELY CITED: REFORMING ACCOUNTABILITY FOR GENERATIVE AI USE BY ADVOCATES AND JUDGES IN INDIA
Authors: RASHAP BANSAL
- Journal IJLRA
- ISSN 2582-6433
- Volume 3
- Issue 7
About Journal
International Journal for Legal Research and Analysis
- AbbreviationIJLRA
- ISSN2582-6433
- AccessOpen Access
- LicenseCC 4.0
All research articles published in International Journal for Legal Research and Analysis are open access and available to read, download and share, subject to proper citation of the original work.
Disclaimer: The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of International Journal for Legal Research and Analysis.