This discussion paper, authored by Jameela Sahiba (Associate Director – AI and Public Affairs, The Dialogue), Sneha Jain (Partner, Saikrishna & Associates), Akshat Agrawal (Senior Associate, Saikrishna & Associates), and Mishthi Dubey (Senior Associate, Saikrishna & Associates), examines the evolving intersection between Generative Artificial Intelligence (GenAI) and copyright law in India.

As GenAI models increasingly rely on large volumes of data—often comprising copyrighted works—for training, this paper interrogates how India’s existing legal framework, particularly the Copyright Act, 1957, applies to such use cases. It focuses on key provisions relating to reproduction and adaptation and assesses how these interact with AI model training. The authors explore fundamental tensions between the rights of content creators and the broader public interest in technological advancement.

This is the first in a two-part series exploring the copyright implications of GenAI. While this paper zeroes in on the legality of inputs—the copyrighted material used to train models—the next will address the far thornier issue of outputs generated by AI systems.

Can existing copyright laws accommodate the realities of AI development? Where should the lines be drawn between fair use, innovation, and infringement?

Download the full paper to explore these critical questions and more.