India court decision on OpenAI copyright

In late July 2026, the Delhi High Court addressed a copyright suit filed by Asian News International (ANI) against OpenAI. On July 24, the court ruled that OpenAI’s use of ANI’s news content to train ChatGPT fell under India’s fair‑dealing exemption for research, rejecting ANI’s request for an interim injunction and noting no proof that the model memorised or reproduced the articles.

A few days later, the same judge clarified the interim decision, emphasizing that the GPT‑4 and GPT‑4o models were trained on data only up to April 2022 and April 2024, while the ANI articles cited were from August and September 2024 and could not have been part of the training set. The court attributed similarities in outputs to Retrieval‑Augmented Generation, which pulls real‑time information, and reaffirmed that facts in news articles are not copyrightable. The ruling constitutes an interim win for OpenAI, while the court indicated that the substantive question of whether RAG‑based outputs constitute a public communication will be addressed in the main proceedings.