Fair Dealing for Machines: The Delhi High Court's First Reading of AI Training Under Copyright Law
UPSC-standard MCQs with explanations, trap analysis, and approach guide. Answer after the test — not before.
1
Easy
1
Medium
1
Hard
Practice this set
3 questions · full analysis after submission · no sign-up required
Article summary
The Delhi High Court has declined to restrain OpenAI in the copyright suit brought by the news agency ANI, holding that the use of copyrighted content to train a generative AI model does not, at the interim stage, constitute infringement. Justice Amit Bansal observed that OpenAI's storage and use of ANI's literary works to train its large language model falls within fair dealing under Section 52(1)(a) of the Copyright Act, 1957, and that ANI had failed to establish a prima facie case that ChatGPT reproduced or retrieved its original works in its outputs. It is the first detailed attempt by an Indian court to locate AI training within the Copyright Act. The order came days after Anthropic's roughly 1.5 billion dollar settlement in a United States authors' class action, producing a striking contrast: on the same underlying question of training AI on copyrighted material, an Indian court declined interim relief while a US case resolved through a large payment.
What this tests
Sample questions — answers revealed after test
Q1. The 'fair dealing' provision under Section 52 of the Copyright Act, 1957, which the Delhi High Court applied, is best described as which of the following?
Q2. The court separated the question of training an AI model on works from the question of whether the model reproduced those works in its outputs. Why is this distinction analytically important?
Q3. Consider the following statements about the Delhi High Court's order: 1. It was an interim order refusing to restrain OpenAI, not a final decision of the suit. 2. The difference in outcome from the United States, where a similar dispute was settled for a large sum, reflects the structural difference between a closed fair dealing list and an open fair use standard. 3. Because the court refused interim relief, it has conclusively held that training AI on copyrighted works is lawful in India. Which of the statements given above are correct?