India Consults on Draft Guidelines for AI Use in Trade Mark Proceedings: Declare AI, Verify Case Law, Face Struck-Off Material and Discipline for Abuse (Not in Force)
India's IP authority, the CGPDTM (Office of the Controller General of Patents, Designs and Trade Marks), published on 17 October a draft set of Guidelines on the Use of Artificial Intelligence in Proceedings before the Trade Marks Registry and opened them for public comment. The draft was prepared on the directions of the Delhi High Court in W.P.(C)-IPD 16/2026 (Stryten Energy LLC v. Nidhi Agarwal & Ors.) and is not yet in force.
The court flagged AI and squatting; this draft is about AI
A clarification worth making: the Delhi High Court's directions and the public-notice title address two things — the use and misuse of AI, and the menace of squatting. But the substantive text now out for comment is titled Guidelines on the Use of AI in Proceedings before the Trade Marks Registry, and chiefly regulates AI. Squatting remains within the court's remit; the document on the table is the AI rulebook.
The core: declare AI use, verify case law and facts yourself
The draft sets out general principles — human primacy and accountability, transparency and disclosure, verification before submission, and integrity of proceedings. In concrete terms: anyone who uses AI to produce substantive legal content in a trade mark filing must make a mandatory declaration of AI use; and any case law or facts generated by AI must be independently verified, with full responsibility resting on the person who files.
Which proceedings it covers
The scope is broad: it would apply to all proceedings pending or hereafter initiated before the Trade Marks Registry across India — applications, replies to examination reports or show-cause notices, oppositions, rectification and cancellation, renewals, assignments and recordals — wherever a document's substantive legal content was prepared with AI's help. The guidelines would operate alongside, not in place of, the Trade Marks Act 1999, the Trade Marks Rules 2017 and the Bharatiya Sakshya Adhiniyam 2023.
What misuse would cost
The draft draws a clear line: fabricated, false or non-existent case law citations, factual misrepresentation, false or misleading declarations, and AI-tampered evidence are all prohibited. Where established, the offending material may be struck off the record or disregarded; the filer may also face disciplinary proceedings — trade mark agents under the applicable Code of Conduct (or, until it is in force, an ad hoc disciplinary committee), and an authorised representative who is an advocate referred to the competent Bar Council. The draft also sets AI duties for examiners and Registry officers and creates a channel for reporting AI hallucinations.
J&C notes that this is still a consultation draft, not yet in force, but the direction is clear: in Indian trade mark work, using AI to draft submissions or search case law is not banned, but whoever files bears the responsibility — every AI-supplied citation and fact must be checked line by line, the verification kept on record, and AI use declared truthfully against the final text. Firms should update their internal AI-use and review workflows accordingly, so that a single hallucinated precedent does not get material struck off or trigger disciplinary proceedings.
Source: ipindia.gov.in | ipindia.gov.in
