OpenAI disputes Indian court's demand to erase ChatGPT data

OpenAI has challenged the jurisdiction of an Indian court to compel the removal of ChatGPT’s training data, asserting that such an order would conflict with its legal obligations under US law, Reuters reported, citing a recent court filing.

The case, filed by Indian news agency ANI in November, alleges that OpenAI used its published content without authorization to train ChatGPT, an AI language model that has gained significant traction in India.

ANI has sought to delete its content from the model’s training dataset, accusing OpenAI of copyright infringement.

OpenAI response

In an 86-page submission to the Delhi High Court on Jan. 10, OpenAI stated that US laws require it to preserve training data while litigation is pending, making data removal orders from India incompatible with its legal responsibilities.

OpenAI also emphasized that the court lacks jurisdiction over its operations, as the company has no physical presence or servers located in India.

The firm stated in the filing:

“The servers on which ChatGPT’s training data are stored are located outside of India.”

While OpenAI has committed to ceasing the use of ANI’s content in the future, ANI contends that the material already incorporated into ChatGPT’s dataset continues to pose risks of unauthorized reproduction.

ANI has also expressed concerns over OpenAI’s commercial agreements with international media outlets, alleging unfair competition. The company contends that ChatGPT has reproduced verbatim or near-verbatim extracts of its articles in response to user prompts.

Court hearing

ANI, which is partially owned by Reuters, has countered that the Delhi High Court holds the authority to address its concerns.

A court hearing on the matter is scheduled for Jan. 28. Reuters clarified that it is not involved in ANI’s business decisions or legal disputes.

The lawsuit is part of a broader wave of copyright challenges against AI firms for allegedly misusing copyrighted material to train models.

In the US, OpenAI is also defending against similar claims, including a high-profile case brought by The New York Times. OpenAI has consistently maintained that its AI systems rely on publicly available data and fall under fair use protections.

Mentioned in this article