The Next Big Copyright Question in India

Generative Artificial Intelligence has created one of the most significant legal crossroads in modern copyright law. Large Language Models (LLMs) are trained using enormous volumes of text, images, music, and other creative works, many of which are protected under copyright law.

Across the world, authors, publishers, artists, and technology companies are engaged in legal disputes over a fundamental question:

Does training an AI model on copyrighted content require permission from copyright owners, or does it fall within statutory copyright exceptions?

In India, the answer remains uncertain.

India's Copyright Framework

The first instinct is to examine Section 52 of the Copyright Act, 1957, which outlines India's fair dealing exceptions.

A key question arises:

Can AI developers argue that training an LLM using copyrighted material amounts to "research" or another protected activity under Section 52?

A closer reading of the law suggests that Section 52, in its current form, does not provide a clear safe harbour for commercial AI training. Expanding the provision beyond its intended scope would require judicial or legislative intervention rather than broad interpretation.

Section 52: A Limited Exception

Unlike the United States, which follows the flexible doctrine of Fair Use, India follows a statutory Fair Dealing model.

Section 52 specifically lists situations where copyrighted works may be used without constituting infringement. These generally include:

  • Private or personal use
  • Research and private study
  • Criticism or review
  • Reporting of current events
  • Educational purposes

This distinction is important because Parliament intentionally created specific exceptions rather than adopting an open-ended balancing test like Fair Use.

As a result, courts cannot simply create new exceptions because technological developments appear beneficial or promote innovation. Instead, they must interpret Section 52 according to its statutory language and legislative intent.

Judicial Interpretation: University of Oxford v. Rameshwari Photocopy

The Delhi High Court's decision in University of Oxford v. Rameshwari Photocopy illustrates this approach.

The Court interpreted Section 52 in accordance with its educational purpose while remaining within the statutory framework enacted by Parliament. It did not suggest that courts possess unlimited authority to create new categories of permissible use simply because technology has evolved.

Can AI Training Be Considered Research?

Supporters of broad AI exceptions argue that training Large Language Models is fundamentally a research exercise.

According to this view, AI systems do not simply reproduce books or articles for users. Instead, they analyse statistical patterns and relationships across millions of works.

However, critics argue that this comparison overlooks an important distinction.

Human researchers typically read copyrighted works. AI developers, however, generally create extensive copies of copyrighted material during the training process. Those copies themselves may raise copyright concerns before any statistical analysis begins.

Furthermore, many modern LLMs are no longer purely research projects. They power subscription-based AI platforms, enterprise software, and commercial products that generate significant revenue.

Given their commercial nature, equating industrial-scale AI training with the "private study" or "research" contemplated under Section 52 becomes increasingly difficult.

The Civic Chandran Decision

The Kerala High Court's decision in Civic Chandran v. Ammini Amma emphasized examining the purpose and character of the impugned use while assessing fair dealing.

Even if AI development involves research components, that alone does not automatically place large-scale commercial AI training within the statutory protections of Section 52.

Why This Debate Matters

The implications extend far beyond AI companies.

The outcome will affect countless creators whose works may become part of AI training datasets without consent or compensation, including:

  • Authors
  • Journalists
  • Publishers
  • Musicians
  • Artists
  • Researchers

The interpretation of Section 52 will therefore shape not only copyright litigation but also the future of India's AI ecosystem.

The Policy Balance

A Broad Interpretation Could:

  • Encourage AI innovation
  • Support Indian startups with limited resources
  • Reduce the cost of AI development
  • Accelerate domestic AI research

A Narrow Interpretation Could:

  • Strengthen copyright protection
  • Ensure creators receive authorization or compensation
  • Promote licensing markets for AI training data
  • Provide greater certainty for rights holders

Conclusion

India now faces a defining legal question: should commercial AI training using copyrighted material fall within the existing framework of Section 52, or should Parliament introduce new statutory provisions specifically addressing artificial intelligence?

The answer will influence not only future copyright disputes but also India's innovation strategy, investment climate, and the balance between technological progress and the rights of creators.

Until legislative reform or a definitive judicial ruling emerges, the legal position surrounding AI training and copyright in India remains uncertain, making it one of the most significant copyright questions of the AI era.

#Artificial Intelligence#AI Governance#European Union#American AI Act#Digital Omnibus#AI Statute
TT
Tunetradr Editorial Verified by Tunetradr — This article has been reviewed, fact-checked and published by our editorial team to ensure accuracy and reliability for our readers.

The Tunetradr editorial team writes practical, no-fluff guides on music distribution, royalties, rights and growing as an independent artist in India.