A coalition of Indian book publishers and their international partners has initiated a copyright infringement lawsuit against OpenAI in the Delhi High Court. The legal action marks the latest in a growing global effort to restrict the use of proprietary content in training artificial intelligence systems like ChatGPT.

The Federation of Indian Publishers (FIP), representing prominent members such as Bloomsbury, Penguin Random House, Cambridge University Press, Pan Macmillan, Rupa Publications, and S. Chand and Co., announced the filing.

Pranav Gupta, the federation’s general secretary, stated that the case centres on ChatGPT’s use of copyrighted book content to generate summaries, potentially impacting sales and creativity within the publishing industry.

“Our demand is clear. OpenAI must cease accessing our copyrighted works. If they refuse to negotiate licensing agreements, they must delete datasets used in AI training and outline how they intend to compensate us. This misuse directly undermines creativity and intellectual property,” said Gupta.

The lawsuit was filed in December 2024 but only came to light recently. OpenAI, known for spearheading the generative AI revolution with its ChatGPT launch in November 2022, has yet to comment on the allegations. The company has consistently maintained that its AI systems adhere to the principle of fair use by utilising publicly available data.

Broad legal implications

The Delhi High Court is already hearing a related lawsuit filed by Indian news agency ANI against OpenAI, making this the country’s most prominent legal challenge in the realm of generative AI and copyright. The publishers’ federation seeks to merge its case with ANI’s, highlighting the broader implications for intellectual property rights and technological progress in India.

“These cases represent a watershed moment,” remarked Siddharth Chandrashekhar, a Mumbai-based lawyer. “The judgments will determine how we strike a balance between protecting intellectual property and fostering innovation in AI.”

In response to the ANI lawsuit, OpenAI argued that any directive to delete training data would breach its legal obligations in the United States. It further contended that Indian courts lack jurisdiction over a copyright case involving servers located abroad. The publishers’ federation, however, countered that OpenAI’s provision of services within India subjects it to Indian laws.

Concerns over book summaries

The heart of the federation’s concern lies in ChatGPT’s ability to provide detailed book summaries. For instance, when asked about the first volume ofHarry Potter by J.K. Rowling, ChatGPT delivered a chapter-by-chapter breakdown but refrained from reproducing the actual text, citing copyright restrictions.

Despite such limitations, Gupta believes the availability of detailed summaries could discourage readers from purchasing books. “Why would anyone buy books if they can access free summaries and extracts through these tools?” he said. “Our members are deeply worried about the potential impact on sales.”

In November 2024, Penguin Random House introduced a global initiative to include a copyright notice explicitly prohibiting the use of its books for AI training.

The road ahead

The federation claims to have “credible evidence” suggesting OpenAI utilised its members’ works for training ChatGPT. The Delhi High Court registrar directed OpenAI to respond to the allegations during a preliminary hearing on 10 January. A full hearing before a judge is scheduled for 28 January.

(With inputs from Reuters)