Delhi High Court's Landmark Ruling on AI and Copyright: A Boost for Innovation

AI's Impact on Information Access

New Delhi: The Delhi High Court recently emphasized that artificial intelligence (AI) has transformed the way the public accesses information. The court expressed concerns that restricting ChatGPT from utilizing copyrighted materials could hinder the advancement of AI and large language models (LLMs) being developed in India.

Justice Amit Bansal highlighted India's leading position in AI, stating that 'data is the oil for large language models to function effectively.'

LLMs are AI models designed to read, comprehend, and generate text.

These remarks were made during an interim ruling related to a lawsuit filed by a news agency against OpenAI, the organization behind ChatGPT.

The court determined that ChatGPT's use of the news agency's copyrighted content constituted fair dealing under the Copyright Act, thus rejecting the agency's request for a temporary injunction against ChatGPT.

The ruling further noted that imposing an injunction at this juncture would negatively affect public interest and the millions of ChatGPT users in India.

'AI, particularly generative AI, has significantly changed how individuals worldwide, including in India, seek and acquire information. Granting an interim injunction now would, in my view, be harmful to the growth of AI and the LLMs being developed in India,' the court stated.

'I am of the initial opinion that OpenAI's practice of using the news agency's original literary works for training LLMs related to ChatGPT falls under Section 52(1)(a) of the Copyright Act, and therefore, does not constitute infringement under Section 51 of the Copyright Act,' the court concluded.

In a detailed 135-page ruling, the court elaborated that training LLMs for ChatGPT contributes to scientific advancement, fosters innovative computational tools, spreads knowledge, and enhances education. While the rights of copyright holders are significant, the societal advantages stemming from such scientific and technological progress are also crucial.

'Thus, based on this preliminary analysis, the court believes that the process of training LLMs for ChatGPT by OpenAI, using the stored literary works of the news agency, qualifies as 'private or personal use, including research' as outlined in Section 52(1)(a) of the Copyright Act and meets the 'purpose' test,' it stated.