Generative AI and Copyright Law in India Reassessing Ownership, Infringement, and Liability SRIHARINI CHELLAPPAN

Generative AI and Copyright Law in India: Reassessing Ownership, Infringement, and Liability

Sriharini Chellappan,

II Yr LLB, IFIM Law School, Bengaluru

Abstract

The rapid spread of generative AI systems is creating complex, unresolved issues in copyright law, where the current laws weren't designed to handle the creation of content by machines. This paper explores the gaps in existing copyright framework in India to handle the risks posed by generative AI. The generative AI models use massive datasets pulled from publicly available sources. This often includes copyrighted materials. Does this use count as "reproduction" under Section 14 of the Copyright Act, 1957 or Can it be considered as fair dealing under Section 52? Indian copyright law doesn't currently have a clear exception for these use cases, leaving AI developers to use these legal gaps. Additionally, this paper explores existing ownership rules and probes if AI generated content to be classified as exclusive ownership, with a new category of protection. Moreover, the possibility of AI outputs closely resembling copyrighted works makes it difficult to apply traditional infringement rules. This is because these models replicate styles or expressions from their training, may not be exact words. Further, the dispersed nature of AI systems which is spread across developers, deployers, and users makes it difficult to establish the accountability structure to protect the fair use. The paper identifies and proposes the specific legislative changes in the Indian Copyright Law that will address accountability during the training stage, regulate the outputs, and provide clearer guidelines for liability.

Keywords: Generative AI, Copyright Act 1957, AI’s Ownership, AI Accountability