Evolving Indian Copyright Law Faces Challenges from Generative AI Innovations

Artificial intelligence (AI) is rapidly transforming multiple facets of daily life and business worldwide. Among its most impactful forms is generative AI (GenAI), which produces extensive content swiftly by leveraging human-generated inputs and deep learning algorithms to identify patterns within training data.

While AI technologies such as chatbots have existed since the 1960s, the public launch of ChatGPT in 2022 marked a turning point, popularizing GenAI tools and accelerating AI innovation. Alongside ChatGPT, tools like Grammarly, DALL·E 3, and Midjourney have become integral to everyday workflows.

This swift technological advancement blurs the distinction between human creativity and machine-generated content, sparking intense debate about the adequacy of existing legal frameworks to govern AI-driven innovation. Intellectual property (IP) law, particularly copyright, is at the forefront of these discussions.

Key concerns focus on rights in AI-generated content, the extent of human involvement required for copyright protection, and the legality of using copyrighted works to train AI models. Some stakeholders argue that AI-created works, if protectable at all, should be treated differently from traditional human-authored works.

This article explores three critical aspects of GenAI’s impact on copyright law: the potential for AI to infringe copyrights, the protectability of AI prompts, and the scope of protection for AI-generated outputs. It also examines foundational copyright principles—originality, authorship, infringement, and fair use—in the context of AI, assessing whether current Indian law can accommodate these rapid technological changes.

A central tension exists: authors risk losing economic value if their copyrights are not respected, yet innovation depends on the ability to use large volumes of copyrighted content to train AI systems.

GenAI systems ingest vast datasets comprising texts, images, sounds, and graphics. Their outputs often combine elements from these datasets, sometimes closely resembling existing works. This raises concerns within creative communities about potential declines in the value of original content.

Legally, the permissibility of using copyrighted materials for AI training varies internationally. The United States employs a ‘fair use’ doctrine allowing limited use without prior consent, though its application to AI training remains ambiguous. India follows a ‘fair dealing’ doctrine under the Copyright Act, 1957, permitting limited use for purposes such as research, criticism, and news reporting. However, India’s fair dealing scope is narrower than fair use, and its applicability to AI training is yet untested and unclear.

Determining copyright infringement hinges on whether AI-generated content is ‘substantially similar’ to protected works and whether it falls within statutory exceptions under section 52 of the Copyright Act. AI-generated literary works used in compliance with these exceptions do not constitute infringement. Conversely, commercial use or use outside these exceptions may infringe copyright.

Another complex issue is whether AI-generated content qualifies as a ‘derivative work.’ Such works are based on existing copyrighted material but include significant modifications. If GenAI produces content that substantially alters original works yet retains identifiable elements, questions arise about infringement liability.

Unlike jurisdictions such as Singapore and Japan, which have explicit provisions permitting the use of copyrighted content for AI training, India lacks clear legal guidelines. Establishing such clarity is essential to balance the interests of creators, AI developers, and the broader innovation ecosystem.

As generative AI continues to evolve, Indian copyright jurisprudence faces the urgent task of adapting to these technological advances. Clear legal frameworks are needed to reconcile the protection of creative works with the facilitation of AI-driven innovation.