Delhi High Court Rules AI Training Does Not Infringe Copyright in ANI-OpenAI Case
The Delhi High Court has held that OpenAI's use of news agency ANI's content to train the AI models behind ChatGPT does not amount to copyright infringement. It is the first substantive judicial finding in India on whether AI companies can use copyrighted news content to train large language models without a licence. ANI had sued OpenAI in November 2024, alleging its published reports were used without permission and that the chatbot generated fabricated stories attributed to ANI. Justice Amit Bansal ruled that ANI had failed to establish that ChatGPT memorised or reproduced its copyrighted reports when answering user queries. The court further held that OpenAI's storage of ANI's articles for training falls under the fair-dealing exemption for research under India's Copyright Act, and therefore does not infringe copyright. The judgment could become an important precedent for India's AI ecosystem, though similar cases continue in the United States, Canada and elsewhere.
Key Facts & Details
10 points- 1The Delhi High Court ruled that AI training on ANI's content does not infringe copyright.
- 2It is the first substantive judicial finding in India on AI training and copyrighted news content.
- 3ANI sued OpenAI in November 2024 over unlicensed use of its reports.
- 4ANI also alleged ChatGPT fabricated stories and attributed them to the agency.
- 5Justice Amit Bansal held ANI failed to show ChatGPT memorised or reproduced its reports.
- 6Storage of the articles for training falls under the fair-dealing exemption for research in the Copyright Act.
- 7The ruling could become an important precedent for India's AI ecosystem.
Deep Dive
- +OpenAI has argued that AI models learn statistical patterns rather than storing individual works.
- +Publishers and authors counter that unlicensed use deprives them of revenue.
- +Similar cases against AI companies continue in the United States and Canada.
Exam Focus
Which court ruled on the ANI-OpenAI copyright case, and on what legal ground was AI training held not to infringe?
Exam Relevance & Angle
This is the first Indian judicial ruling on AI training and copyright, and the court, judge and fair-dealing ground are all directly testable.
Target Exams
Background & Context
Copyright gives an author control over copying their work, not over anyone learning from it. The legal question with AI is which of the two is happening. AI developers argue a model absorbs statistical patterns across millions of texts and stores no particular article, so nothing is copied in the legal sense. Rights holders argue the training set itself is a copy, made without payment. India's Copyright Act contains a fair dealing exception permitting use of a work for purposes including research — a safety valve so that study and analysis are not blocked by copyright. The Delhi High Court read AI training into that exception, and separately found no evidence the model was reproducing ANI's text back to users.
Test Yourself
1 / 2On what ground did the Delhi High Court hold that OpenAI's storage of ANI's articles for AI training was lawful?
Source
This topic is important for:
