What happened: a legal explainer on text-and-data for AI under Indian copyright
A legal explainer panel, linked to the broader AI-books story in India, describes how Indian copyright doctrines and exceptions are understood when AI systems copy text to perform analysis and data processing. The explainer focuses on practical constraints: it outlines when copying may be permissible, what limitations typically apply, and how courts are commonly understood to assess different kinds of AI use.
Background and earlier position: copyright limits vs analysis and processing
Copyright law protects creative works from unauthorized copying, but copyright also contains exceptions and doctrines that can allow limited copying in specific contexts. In the AI text-and-data debate, the key question is how far “copying for analysis and processing” is treated as an allowed activity versus an infringing use.
The explainer panel presents a distinction often used in copyright reasoning: non-commercial research and transformative use (use that changes the purpose and character of the material) may receive more favourable treatment, while copying aimed at building commercial datasets may face stronger resistance. The panel also highlights that courts’ approach depends on the purpose and the scope of copying, even when the copying is technically necessary for AI training or evaluation.