The Delhi High Court has delivered a landmark ruling in favor of OpenAI, stating that the company’s use of copyrighted material to train its artificial intelligence models does not violate copyright law. The decision, announced on Friday, rejected an interim injunction sought by the news agency ANI in its ongoing copyright lawsuit against the AI developer.
Justice Amit Bansal held that OpenAI’s act of storing ANI’s works for training its Large Language Models (LLMs) does not, by itself, constitute copyright infringement. The court further found that ANI failed to establish a prima facie case that ChatGPT had memorized or reproduced its copyrighted content in its responses.
“OpenAI’s act of storing ANI’s works does not amount to copyright infringement,” Justice Bansal observed, adding that “ANI had failed to show that any memorization or regurgitation of ANI’s work has happened from responses generated by ChatGPT.”
‘Irreparable Injury’ to Public if AI Training is Stopped
The court emphasized that granting an interim injunction would have wider consequences beyond the parties involved. Justice Bansal highlighted that irreparable injury would be caused not only to OpenAI but also to the public at large if such an injunction were granted. The ruling noted that restricting AI companies from training their models using copyrighted material could adversely affect technological innovation and scientific progress.
ANI Sought Rs. 2 Crore in Damages
ANI filed the lawsuit in November 2024, alleging that OpenAI had used its copyrighted news reports without permission to train ChatGPT. The agency sought Rs. 2 crore in damages and asked the court to permanently restrain the company from storing, publishing, reproducing, or using its content. ANI argued that ChatGPT generated responses that were “verbatim or substantially similar” to its reports, and also alleged that the chatbot produced “hallucinated” responses falsely attributed to the agency.
OpenAI Cited Publicly Available Data
OpenAI rejected the claims, stating that its AI models are trained based on publicly available data. According to the court, OpenAI informed the court that it had blacklisted the domain name of ANI in October 2024, following which no new content from the website can be used for training any AI model. Senior counsel Amit Sibal, representing OpenAI, argued that “copyright protects the expression of ideas; there is no monopoly on facts” and noted that OpenAI’s servers are located outside India.
This case represents one of the first landmark decisions in India concerning the use of copyrighted works for training generative AI, setting a significant precedent for the technology sector.