ChatGPT: India media seek to join lawsuit against OpenAI chatbot

Several major Indian news organizations are seeking to join a lawsuit against OpenAI, the American startup behind ChatGPT, alleging unauthorized use of their content. These news organizations include some of India's oldest publications, such as The Indian Express, The Hindu, India Today Group, and NDTV, owned by billionaire Gautam Adani, as well as over a dozen other media outlets.

OpenAI has denied these allegations, informing the BBC that it uses "publicly available data" in accordance with "widely accepted legal precedents." OpenAI CEO Sam Altman discussed plans for a low-cost artificial intelligence ecosystem in India with Indian Minister of Information Technology Ashwini Vaishnaw in New Delhi on Wednesday.

Altman stated that India "should be one of the leaders of the AI revolution," and claimed that his earlier comments in 2023 regarding the difficulty for Indian companies to compete were taken out of context. He told local media at the event: "India is an extremely important market for the entire field of AI, and especially for OpenAI."

The lawsuit filed by India's largest news agency, Asian News International (ANI), against OpenAI in November last year is the first such case in India. ANI accuses ChatGPT of illegally using its copyrighted material—a charge OpenAI denies—and is seeking 20 million rupees ($230,000; £185,000) in damages. This case is significant given ChatGPT's planned expansion in the country. A survey indicates that India already has one of ChatGPT's largest user bases. Chatbots like ChatGPT are trained on vast datasets collected by scraping the internet. Content generated by India's nearly 450 news channels and 17,000 newspapers provides enormous potential for this purpose. However, it remains unclear what materials ChatGPT can legally collect and use for this purpose.

OpenAI faces at least a dozen lawsuits globally from publishers, artists, and news organizations, all accusing ChatGPT of using their content without permission. The most notable of these is the lawsuit filed by The New York Times in December 2023, which seeks "billions of dollars" in damages from OpenAI and its backer, Microsoft. Vibhav Mithal, a lawyer specializing in artificial intelligence at the Indian law firm Anand and Anand, stated: "Any court ruling will have some persuasive value for other similar cases around the world." Mithal suggested that the verdict in the ANI lawsuit could "define how these AI models will function in the future" and "what copyrighted news content can be used to train AI generative models [like ChatGPT]."

What is ANI's case? ANI provides news to its paying subscribers and owns exclusive copyrights to a vast amount of text, images, and videos. In its lawsuit filed with the Delhi High Court, ANI states that OpenAI used its content to train ChatGPT without permission. ANI argues that this has made the chatbot better and enabled OpenAI to profit. The news agency stated that before filing the lawsuit, it informed OpenAI that its content was being used illegally and offered to grant the company a license to use its data. ANI says that OpenAI rejected this offer and placed the news agency on an internal blocklist to prevent its data from being collected further. It also asked ANI to disable certain web crawlers to ensure that its content was not scraped by ChatGPT. The news agency says that despite these measures, ChatGPT still scraped its content from its users' websites. It says that this "unjustly" enriched OpenAI.

ANI also stated in the lawsuit that the chatbot generates its content verbatim in response to certain prompts. ANI stated that in some instances, ChatGPT incorrectly attributed statements to the news agency, damaging its credibility and misleading the public. In addition to seeking damages, ANI is also asking the court to direct OpenAI to stop storing and using its works. In its response, OpenAI stated that it opposes the lawsuit being filed in India because the company and its servers are not located in the country, and the chatbot was not trained there.

News organizations seek to join the lawsuit. In December, the Federation of Indian Publishers, claiming to represent 80% of Indian publishers, including Penguin Random House and Oxford University Press India, filed an application with the court stating that they are "directly affected" by the case and should be allowed to present their arguments. A month later, the Digital News Publishers Association (DNPA), representing major Indian news media, and three other media organizations also filed similar applications. They argued that while OpenAI has reached licensing agreements with international news publishers such as the Associated Press and the Financial Times, it has not adopted a similar model in India. The DNPA told the court that the case will affect the livelihoods of journalists and the entire news industry in the country. However, OpenAI argued that the chatbot is not a "substitute" for news subscriptions and is not used for such purposes.

The court has not yet admitted these applications from the publishers, and OpenAI argues that the court should not hear them. But the judge clarified that even if these associations are allowed to argue, the court will be limited to ANI's claims because the other parties have not filed their own lawsuits. Meanwhile, OpenAI told the BBC that it is engaged in "constructive partnerships and conversations" with news organizations around the world, including in India, to "work together."

What is the current state of AI regulation in India? Analysts say that lawsuits against ChatGPT around the world may bring attention to certain aspects of the chatbot that have not yet been scrutinized. Dr. Sivaramakrishnan R Guruvayur, whose research focuses on the responsible use of artificial intelligence, said that the data used to train the chatbot is one such aspect. He said that the ANI-OpenAI case will lead the court to "assess" the chatbot's "data provenance."

Governments around the world have been struggling to figure out how to regulate artificial intelligence. In 2023, Italy blocked ChatGPT, stating that the chatbot's large-scale collection and storage of personal data raised privacy concerns. The European Union approved a law last year to regulate artificial intelligence. The Indian government has also stated its intention to regulate artificial intelligence. Ahead of the 2024 general election, the government issued a recommendation that AI tools that are "under testing" or "unreliable" should obtain government permission before being launched. It also required AI tools not to generate responses that are illegal in India or "threaten the integrity of the electoral process."