ChatGPT sued for copyright infringement, OpenAI gets Delhi High Court's summons

The Delhi High Court issued a notice to OpenAI, seeking its response to a request for temporary relief made by news agency ANI.

In Short

• News agency ANI alleges ChatGPT uses its copyrighted content
• OpenAI rebuts claims, says it didn't violate any copyright rules
• Delhi High Court lists matter for hearing on January 28 next year

The Delhi High Court has taken up the first-ever lawsuit in India against OpenAI after news agency ANI alleged copyright infringement by its AI tool, ChatGPT. On Tuesday, the court decided to appoint a legal expert, known as an amicus curiae, to assist with the case due to its complex nature and the broader implications of AI advancements on copyright laws.

The court issued a notice to OpenAI, seeking its response to a request for temporary relief made by the news agency.

However, Justice Amit Bansal declined to issue an interim injunction, stating, “It’s a complex matter which requires deeper examination.”

Advocate Sidhant Kumar, representing ANI, argued that OpenAI had used the agency’s copyrighted material to train ChatGPT. He explained that the agency provides certain news items exclusively to its subscribers, which are not intended for the public. However, these materials allegedly appear as part of ChatGPT’s responses.

Kumar also claimed that ChatGPT sometimes attributes false statements to the agency, spreading misinformation. He said, “It hurts not just my private rights, but also spread fake news and public disorder. There’s also a public angle to this.”

Senior Advocate Amit Sibal, appearing for OpenAI, questioned the Delhi High Court’s jurisdiction over the case. He argued that OpenAI operates internationally, with no servers or offices in India, and stated, “The plaintiff has no cause of action against us in India.”

Sibal defended OpenAI and said that ChatGPT does not store or reproduce specific content from the news agency’s database, copyright law protects specific expressions, not ideas or facts, and news forms a tiny portion of the data used by ChatGPT, and the plaintiff’s contribution is even smaller.

He also pointed out that OpenAI has measures in place for website owners to block their content from being accessed. “We’ve been transparent and bona fide from the start,” Sibal said, adding that OpenAI’s disclaimers acknowledge that its AI is not perfect and is continually improving.

The court expressed particular concern about allegations that ChatGPT spreads false information.

While OpenAI’s counsel refuted these claims, asserting that such instances are rare, the plaintiff’s counsel argued that the platform’s inaccuracies undermine trust and divert traffic away from the agency’s website.

Why would anyone pay for a subscription when the data is freely available on ChatGPT, Kumar asked.

The court has directed both parties to file detailed responses, while the amicus curiae will submit an independent report. The case is scheduled for its next hearing on January 28, 2025.