Context: Last November, India’s Asian News International (ANI) sued OpenAI in the Delhi High Court over copyright infringement, seeking a preliminary injunction (November 18, 2024 ai fray article). ANI became the first in India – albeit the latest of several international media and news organisations – to sue OpenAI for using its copyright-protected content to train ChatGPT.
What’s new: The New Delhi-based Federation of Indian Publishers (FIP) has filed an impleadment application as a party in ANI’s suit. OpenAI used its members’ literary works to train its ChatGPT service, it has also alleged. The application was filed on behalf of all the FIP’s members, including international publishers Bloomsbury, Penguin Random House, Cambridge University Press, and Pan Macmillan.
Direct impact: The FIP’s request follows a number of other actions filed before the Delhi High Court that deal with the issue of AI and deepfake copyright violations. In December, Indian advertising firm Mash Audio Visuals Pvt Ltd., for example, filed a petition seeking the prohibition and punishment of the sale of AI-generated images that are created using the original works of artists without their permission. In that case, the plaintiffs are also seeking major amendments to India’s Copyright Act 1957 (December 5, 2024 ai fray article).
Wider ramifications: Pressure is mounting on OpenAI in courts across the world. The ANI case is one of the latest in a series brought against OpenAI for copyright infringement by media, news and publishing organisations – and marks yet another dispute filed outside the U.S.
This is the court order:
The FIP is the country’s largest national publishing body, representing over 80% of India’s publishing industry. In its request, the federation has argued that it has obtained “credible evidence” from its members that OpenAI used their literary works to train its ChatGPT service.
In a statement, the FIP’s general secretary Pranav Gupta stressed that ChatGPT – a free tool – is essentially producing book summaries and extracts based on copyright-protected content. He said:
“Why would people buy books then? This will impact our sales – all members are concerned about this.”
Mr. Gupta warned that if OpenAI does not want to engage in licensing with its members, then it should delete datasets used in AI training and explain how they will be compensated. “This impacts creativity,” he added.
OpenAI’s growing docket
The FIP’s request comes just before the ANI case was scheduled to be heard in the Delhi High Court (tomorrow). OpenAI said last week that any order to delete training data would be in breach of its legal obligations in the U.S. and that Indian judges have no right to hear a copyright case against OpenAI because its servers are located outside India. But the federation has noted that OpenAI offers services in India so its activities thus fall under Indian jurisdiction.
While India is an important market for the AI service provider – as it is home to over 1.4 billion people – OpenAI is facing dozens of cases in several other major jurisdictions, too. In the publishing industry, it was sued by the Authors Guild and at least 28 different authors in the Southern District of New York – the cases were later consolidated into a single consolidated class-action complaint last year and the plaintiffs are being represented by Susman Godfrey’s Justin Nelson (February 5, 2024 class action complaint). Mr. Nelson is also acting on behalf of book authors in another major publishing copyright infringement class action against Claude LLM maker Anthropic (August 20, 2024 ai fray article).
OpenAI is also facing cases by a handful of news organisations. The New York Times was one of the first to sue it for infringing its copyright-protected articles to train ChatGPT (December 27, 2023 ai fray article). Meanwhile, outside the U.S., OpenAI was hit by a copyright infringement claim in Canada in December. Seven of Canada’s leading news organizations, including the Toronto Star and Metroland Media, have sued it in the Ontario Superior Court of Justice, seeking an injunction and claiming CAD20,000 in statutory damages for every article that OpenAI has allegedly used to train its ChatGPT software unlawfully (December 3, 2024 ai fray article).
In Germany, there have been two high-profile cases brought by German music rights collecting society GEMA over AI-related copyright infringement – although only one is against Open AI and it concerns song lyrics (November 14, 2024 ai fray article). The other GEMA case, filed earlier this month, was brought against Suno – an AI service provider that generates music from its internal repertoire – and asserts copyrighted musical compositions (January 21, 2025 ai fray article).