Delhi High Court upholds OpenAI's use of ANI news for ChatGPT training

The Delhi High Court, hearing a petition by news agency Asian News International (ANI), ruled that OpenAI’s use of ANI’s published news content to train its ChatGPT service does not infringe copyright. Justice Amit Bansal said ANI failed to demonstrate that the AI model memorised or reproduced its articles, and that the use falls under the fair‑dealing exemption for research in India’s Copyright Act.

The court refused to grant an interim injunction, noting that such a restriction would harm both OpenAI and the public interest. This decision marks the first substantive Indian judgment on whether AI companies can train large language models on copyrighted news material without a licence, a question also being litigated in the United States and Canada.