The Delhi High Court has declined, for now, to block OpenAI from using ANI material to train ChatGPT. In an order by Justice Amit Bansal, the court indicated that the use in question currently appears to fall within fair dealing for research under copyright law.
The case centers on whether training an AI system on news reports amounts to copyright infringement. Based on the court’s present view, that threshold was not met at this stage. The judge also noted that there was no evidence before the court showing that ChatGPT had reproduced ANI’s reports in a way that established copying of protected content.
The ruling is significant because it offers an early judicial view in India on a major legal issue facing AI developers and publishers: whether copyrighted material can be used in model training without violating the rights of the original creator. By refusing interim relief, the court has allowed OpenAI’s current practice to continue while the broader dispute remains unresolved.
For the technology and media sectors, the order highlights the growing tension between AI development and copyright protection. It also suggests that Indian courts may examine AI training separately from direct reproduction, especially where the use is argued to support research and there is no clear showing that the original reports were output back to users.
