AI-Generated Content and Copyright in India: Who Owns It?

Two questions about AI and copyright are being asked simultaneously in India right now, and they are frequently confused with each other. The first is whether a business that uses AI to generate content- a marketing copy, a product description, a logo design- owns copyright in that output. The second is whether an AI company that trains its model on existing content is infringing the copyright of the people who created it.

Both questions matter for different reasons. The first affects every business using AI tools in its operations today. The second affects every content creator, publisher, and media organisation whose work is being ingested by AI systems. India’s Copyright Act, 1957 was written decades before generative AI existed India’s Copyright Act, 1957 was written decades before generative AI existed; it is not silent on technology-assisted creation. Under Section 2(d)(vi), introduced in 1994, the author of a computer-generated work is defined as 'the person who causes the work to be created.' However, the Act leaves an open question for Indian courts: how much prompting or creative direction is legally required for a human user to qualify as the person who 'caused' the AI output? It does not answer the question directly; however, what it does provide is a framework that courts are now beginning to apply to situations its drafters never anticipated.

The Authorship Problem: Can AI Own Copyright?

Indian copyright law requires a human author. Section 13 of the Copyright Act, 1957 provides that copyright subsists in original literary, dramatic, musical, and artistic works. The word “original” under Indian law requires that the work originate from a human author who has applied some degree of skill, judgment, and creative effort to produce it.

An AI system does not have legal personality under Indian law. It cannot own property, enter contracts, or hold rights. This means that a work produced entirely by an AI system, with no meaningful human creative input, does not qualify for copyright protection under the current framework. Because an AI system lacks legal personality, it cannot be an author. And as there is no author to vest the rights in, no copyright arises.

This is the straightforward case. The harder cases are the ones businesses actually face day to day.

AI-Assisted Content: Where Human Authorship Still Exists

Most AI-generated content in commercial use is not purely machine-generated. A person writes a detailed prompt, selects from multiple outputs, edits the result, combines it with other material, or uses AI as a tool within a larger creative process. In these situations, the question is whether the human contribution is sufficient to constitute authorship under Indian law.

Indian courts have not yet directly addressed this question for generative AI outputs. But the principle from existing copyright jurisprudence is relatively clear: what matters is whether a human being has exercised sufficient skill and judgment in creating the final work. A person who uses AI to generate a first draft and then substantially edits, curates, and adapts that output is likely exercising the kind of creative judgment that Indian copyright law recognises. A person who types a one-word prompt and accepts the output verbatim is likely not.

The practical consequence for businesses is that the more human creative input goes into an AI-assisted work, the stronger the copyright claim. Documenting that creative process- what instructions were given, what choices were made, what edits were applied- is increasingly important for businesses that want to establish ownership over AI-assisted creative output.

The Training Data Question: ANI vs OpenAI

While the authorship question remains largely untested in Indian courts, the question of whether AI training on copyrighted content constitutes infringement reached the Delhi High Court in a landmark interim ruling delivered on 24 July 2026.

ANI Media Pvt Ltd, one of India’s largest news agencies, sued OpenAI alleging two things: that OpenAI had unlawfully stored and used ANI’s news content to train the large language models behind ChatGPT, and that ChatGPT had generated hallucinated statements falsely attributed to ANI that damaged its reputation.

Justice Amit Bansal delivered a 135-page interim order declining to grant ANI an injunction against OpenAI. On the training claim, the court held, at the prima facie stage, that OpenAI’s storage and use of ANI’s content to train its models fell within the fair dealing exception under Section 52(1)(a) of the Copyright Act, 1957. The court also found that ChatGPT’s outputs were not substantially similar to ANI’s original reporting, and that ANI had not shown that the model had memorised or reproduced its works in user-facing responses.

Two important caveats apply to this ruling. First, it is an interim order, not a final judgment. The main suit continues on merits, and the court explicitly noted that its prima facie findings at the interim stage do not determine the outcome of the full trial. Second, the ruling on the reputation and hallucination claim went in ANI’s favour: the court found it had jurisdiction and that OpenAI’s generation of false statements attributed to ANI was a legitimate concern that the main suit would need to address.

The ANI ruling does not settle the law on AI training and copyright in India. What it does is establish that Indian courts are willing to apply the fair dealing exception to AI training in appropriate circumstances, and that the key factual question is whether the AI output reproduces or substantially resembles the original copyrighted work.

What This Means for Content Creators and Businesses

For content creators and publishers, the ANI ruling signals that preventing AI companies from training on publicly available content through copyright claims alone is a difficult argument in India, at least at the interim stage. The more promising avenue, based on the ruling, may be claims related to AI-generated outputs that reproduce, closely paraphrase, or are substantially similar to the original work, since the court’s finding on the training claim rested partly on the absence of such reproduction in ChatGPT’s outputs.

For businesses using AI tools, the ruling does not create a licence to use any content for AI training. It applies the fair dealing exception to a specific set of facts, and the parameters of that exception- what purposes it covers, how transformative the use must be, and what harm to the copyright owner is acceptable- remain to be worked out in the full trial and subsequent cases.

For businesses commissioning AI-generated creative work, the authorship gap in Indian law creates a practical risk: content generated entirely by AI may not be protectable by copyright, which means a competitor could freely copy it. The response to this risk is to ensure that human creative input is embedded in the production process and documented, so that the copyright claim rests on that human contribution rather than on the AI output alone.

Where the Law Is Headed

India does not yet have legislation specifically governing AI and copyright. The Copyright Act, 1957 is being applied to circumstances it was not designed for, and the ANI case is the first of what will likely be many judicial interventions before a settled framework emerges. The main suit in ANI vs OpenAI continues, and its final outcome will carry significantly more weight than the interim order.

What is clear from the current position is that businesses operating in AI-adjacent spaces, whether as content creators, AI tool users, or AI developers, are operating under genuine legal uncertainty. The decisions being made now about how AI tools are used, what content is ingested, and how creative output is produced and documented will determine the strength of any IP position those businesses hold when the law eventually settles.

M&P IP Protectors is an IPR law firm advising technology companies, media businesses, and content creators on copyright protection and IP strategy in India. To assess how AI tools affect your copyright position or discuss protection for AI-assisted creative work, reach out to our attorneys.