India: Delhi High Court Rules on AI Training and Copyright

On 24 July 2026, the Delhi High Court issued its decision in CS(COMM) 1028/2024, concerning the use of copyrighted content to train AI models and the reproduction of such content through AI-generated outputs. 

The Court first examined whether Indian courts had territorial jurisdiction where OpenAI’s servers and data processing were located in the United States, and whether Indian copyright law could apply to training activities carried out outside India. The Court held that it had jurisdiction under the Code of Civil Procedure and the Copyright Act, noting, among other factors, that the claimant’s principal place of business was in Delhi, OpenAI provided its services to users in India, and relevant outputs were generated within the jurisdiction.

 It further held that the Indian Copyright Act was applicable to the alleged acts, considering the activities in India to form part of a chain leading to the training process, and that this did not amount to an extraterritorial application of Indian law. 

On the substantive copyright issues, the Court found that the claimed reproductions in ChatGPT outputs were not shown to result from the training process and that the storage of copyrighted works for AI training fell within the Section 52(1)(a) exception, concluding that the use in question did not constitute copyright infringement.