OpenAI has told an Indian court that it cannot delete the data used to train its AI model, ChatGPT, as doing so would conflict with its legal obligations in the United States. According to a report by Reuters, the company also argued that Indian courts do not have jurisdiction over the matter because OpenAI has no physical presence or servers in India.
This development comes as part of a lawsuit filed by ANI, a leading Indian news agency, accusing OpenAI of copyright infringement. ANI claims that OpenAI used its articles without permission to train ChatGPT and has demanded that the data be deleted.
In its 86-page court filing dated 10 January, OpenAI explained that it is bound by U.S. laws to preserve its training data due to ongoing copyright disputes in the United States, including a high-profile case brought by The New York Times. The company said it cannot selectively delete ANI’s data while complying with U.S. legal requirements.
OpenAI further argued that it does not operate in India, stating, “The company has no office or permanent establishment in India, and its servers are located outside the country.” Therefore, it believes Indian courts lack the authority to enforce ANI’s demands.
ANI’s claims and concerns
ANI alleges that ChatGPT reproduces parts of its articles in response to user prompts, which it says harms its business and amounts to unfair competition. ANI also raised concerns about OpenAI’s commercial agreements with global media outlets like Time Magazine and The Financial Times, claiming this creates an uneven playing field.
In response to ANI’s accusations, OpenAI denied wrongdoing and argued that ANI had used its own articles as prompts to try and generate similar outputs, calling this an attempt to “manipulate ChatGPT”.
This case is part of a growing trend of copyright disputes against AI companies like OpenAI, as creators of content push back against the use of their work in training AI models. While OpenAI has begun signing licensing deals with some media organisations, questions about compensation and consent remain contentious.
The Delhi High Court will hear the case on 28 January. The outcome could influence how India addresses copyright issues related to AI and whether companies operating from abroad can be held accountable for alleged violations involving Indian content.