AI and Copyright in India — Who Owns Content Created by Artificial Intelligence?

Artificial intelligence is generating text, images, music, code and designs at scale. Businesses are using AI tools to produce marketing content, legal documents, product designs, software and creative works. This raises a question that Indian copyright law has not yet fully answered: who owns the copyright in content generated by an AI system? This guide explores the current legal position in India, the uncertainties that remain and the practical steps creators and businesses should take.

The Current Position Under Indian Law

The Copyright Act, 1957 protects "original literary, dramatic, musical and artistic works." The word "original" has been interpreted by Indian courts to require that the work originate from a human author — that it involves the application of human skill, labour and judgement.

This creates an immediate problem for AI-generated content: an AI system is not a human. It cannot be an "author" under the Copyright Act. This means that content generated entirely by an AI system — without significant human creative input — may not qualify for copyright protection at all, because it has no human author to own the rights.

The Human Creative Input Test

The critical question for AI-generated content in India is: how much human creative input went into the work?

• AI as a tool (significant human input): If a human uses an AI tool to assist in creating a work — providing detailed prompts, selecting and editing AI outputs, combining AI suggestions with original human expression, making creative decisions throughout — the resulting work likely qualifies for copyright protection. The human who made the creative choices is the author.
• AI as autonomous creator (minimal human input): If a person simply types a one-line prompt and publishes whatever the AI outputs without any selection, editing or creative contribution, the copyright position is far less clear. Indian law has not yet definitively addressed this scenario.

The more human creative judgement and selection is involved, the stronger the copyright claim. The more the AI operates autonomously, the weaker the claim.

The "Computer-Generated Work" Provision

Section 2(d)(vi) of the Copyright Act defines the author of a "computer-generated work" as "the person who causes the work to be created." This provision was introduced in 1994, before modern generative AI existed, and was intended to cover works produced by computers following explicit human programming.

Whether this provision applies to works generated by large language models or image generation AI — where the AI's output is not predictable from the input in the way that a traditional computer programme's output is — is debated among Indian IP scholars and practitioners. Some argue this provision covers AI-generated works; others argue it cannot be stretched to cover autonomous generative AI. There is no binding Indian court decision on this question yet.

Who Owns the Training Data Problem

AI systems are trained on vast datasets of existing human-created content — text, images, music and code that may be protected by copyright. This raises a separate question: does training an AI on copyrighted content infringe the copyright in that training data?

This is one of the most actively litigated IP questions globally in 2024–25. In India, the position is currently uncertain. The Copyright Act has a fair dealing exception for research and criticism, but whether commercial AI training qualifies as "research" under this exception has not been tested before Indian courts. The outcome of cases like the ANI v. OpenAI matter — the first major AI copyright case to be filed in India — will be significant in shaping the Indian position.

Practical Steps for Businesses Using AI

Given the current legal uncertainty, here is what businesses and creators using AI tools should do:

• Document human creative input: Keep records of your prompts, the selections you made from AI outputs, the edits you applied and the creative decisions you took. This documentation strengthens your copyright claim in any dispute.
• Do not publish AI outputs unedited: Add human creative contribution — edit, select, combine, arrange — before publishing AI-generated content. This creates a stronger basis for copyright protection.
• Check AI tool terms of service: Some AI tools claim ownership of outputs generated using their platform. Others grant the user full rights. Read the terms carefully.
• Use AI-assisted copyright registration: For significant works created with AI assistance, register the copyright. In the registration application, describe the human creative contributions clearly.
• Protect your training data: If you have proprietary datasets, protect access to them through NDAs, contracts and database protections where available.

What the Future May Hold

India's copyright law is under review. The Draft Digital IP Protection Bill (2025) targets gaps in digital IP enforcement and addresses emerging technologies like AI, signalling that legislative changes are coming. It is likely that India will introduce specific provisions addressing AI-generated works — either extending copyright protection under certain conditions or definitively excluding AI output from protection.

Businesses building on AI-generated content should monitor these developments closely. The legal framework is evolving rapidly, and what is uncertain today may be clarified within the next few years through legislation or court decisions.

Conclusion

The intersection of AI and copyright in India is genuinely unsettled legal territory. The safest approach for creators and businesses is to ensure meaningful human creative input in all AI-assisted works, document that input carefully, and stay updated on legislative and judicial developments. If you are building a business that relies heavily on AI-generated content and need clarity on your IP position, a consultation with an IP specialist is advisable. Learn about copyright registration services → or read more on the IP Law Blog.