
The Delhi High Court denied ANI an interim injunction against OpenAI, ruling that training on public data may be fair dealing. The case highlights unresolved tensions between AI development and copyright.
The Delhi High Court denied ANI Media an interim injunction against OpenAI. The court ruled that training generative AI models on publicly available content could qualify as fair dealing. The decision is preliminary but, according to the Mint article co-authored by Cyril Amarchand Mangaldas lawyers, signals how Indian courts may balance copyright protection with AI development.
The court held that storing publicly available works to train large language models could be fair dealing for private use, including research. Training data retained in a closed environment and not disclosed to the public amounts to private use, the court reasoned. Machine learning, it said, could qualify as research under an updated interpretation of older law.
The court refused to treat OpenAI's commercial character as an automatic disqualification. Research does not cease to be research merely because a company undertakes it or because the resulting product is monetized, the court wrote. That does not mean every commercial use is fair. An activity otherwise covered by fair dealing does not lose that protection solely because it is commercial.
On the question of infringement, the court found insufficient evidence that ChatGPT had memorized or substantially reproduced ANI's protected expression. It distinguished the material used to train a model from the responses generated for users. Infringement at the output stage must be established independently, the court said.
The court's assessment of the competing consequences of an injunction carried its strongest policy signal. ANI had offered OpenAI a licence for $7.5 million, demonstrating that its asserted injury was capable of monetary valuation. If ANI ultimately succeeds, its loss may be compensated. The possible consequences of an injunction were harder to contain. The court observed that requiring licences from multiple sources could make LLM development economically unviable. An injunction could prejudice OpenAI and the public at large, the court concluded.
Publishers' concerns remain legitimate. Journalism requires sustained investment. If AI products reproduce or substitute for publishers' content, they could erode traffic and licensing revenues. The court's separation of training from outputs is critical. Protection for training should not immunize memorization or real-time retrieval that substitutes the original work. The Mint article suggests the next disputes may focus less on what a model learned and more on what an AI product retrieves and monetizes.
The Mint article notes that the judicial and regulatory outlooks may now be moving in different directions. The court found an interim answer within the existing fair dealing framework. The Department for Promotion of Industry and Internal Trade working paper proposes a mandatory licensing framework. Developers would train on content while rights holders receive recurring remuneration through a centralized mechanism. Both approaches seek to enhance AI development. They differ fundamentally on whether rights holders must be paid for enabling such development, the lawyers wrote.
The interim order provides the first judicial answer. It is unlikely to be the last word, the Mint article concludes. The legal position may continue to develop through further judicial and legislative engagement.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.