Deleting ChatGPT Data Violates US Laws: OpenAI Tells Delhi High Court

OpenAI, the company behind ChatGPT, has informed the Delhi High Court that removing training data used for its AI models would violate U.S. legal rules. The company is responding to a lawsuit filed by ANI, a news agency, accusing OpenAI of using its published content without permission to train ChatGPT. ANI filed the case in November 2024, demanding the removal of its content already stored in ChatGPT.

In a detailed 86-page filing submitted on January 10, 2025, OpenAI stated that it could not delete ANI's data because it is facing similar lawsuits in the U.S., where laws require the company to retain the data during ongoing legal proceedings. OpenAI also argued that Indian courts do not have jurisdiction over the matter, as the company has no operations in India.

ANI, in which Reuters holds a 26% stake, disagrees with OpenAI's position. The news agency maintains that the Delhi court has jurisdiction to rule on the case and plans to file a detailed response. ANI’s complaint is that its content remains stored in ChatGPT’s memory, despite OpenAI’s claim that it will no longer use ANI’s material for training.

This lawsuit is part of a growing trend, as OpenAI faces similar legal challenges, including one from the New York Times in the U.S., over accusations of using copyrighted content to train AI models without permission. OpenAI has denied these claims, arguing that its systems only use publicly available data and make fair use of it.

In its January filing, OpenAI reiterated its stance that it will not use ANI’s content going forward, but that the data already within the system cannot be deleted due to U.S. legal requirements. OpenAI has also been transitioning from a non-profit to a for-profit business, securing $6.6 billion in funding last year to stay competitive in the rapidly growing and expensive AI sector.