Free speech includes right to receive information: Broadband India Forum sides with ChatGPT against ANI in Delhi HC

Broadband India Forum, a think tank focused on developing the broadband ecosystem, told the Delhi High Court on Friday that freedom of speech and expression under Article 19 of the Constitution also includes the right to receive information, and Artificial Intelligence (AI) tools like ChatGPT cannot be restrained from using media reports/ original data to answer user queries.

The submission was made by Senior Advocate Kapil Sibal before Justice Amit Bansal during the hearing of Asian News International (ANI)'s copyright infringement suit against OpenAI.

Opposing grant of any interim injunction against ChatGPT, Sibal said that people have a fundamental right to the best means of imparting and receiving information.

"The right to receive information is a species of the right of freedom of speech, an expression guaranteed under Article 19 , a citizen has a fundamental right to use the best means of imparting and receiving information. It is the citizens right that is being violated. You say that 'no you can't access what I put in public domain without any restrictions'," Sibal, who appeared for the think tank said.

ANI's suit has alleged that its original content is being exploited by OpenAI for commercial gain and to train ChatGPT for answering user queries. In response, OpenAI has said that copyright protection in news reporting is very narrow since there exists a greater public interest in dissemination of information.

The Court has earlier heard many intervenors in the case, including the Digital News Publishers Association (DNPA) which has accused OpenAI of infringing the rights of media organisations by training ChatGPT on the basis of online news reports.

Today, Sibal submitted that application seeking interim injunction against ChatGPT cannot be allowed as facts about the working of Large Language Models (LLM) are not there before the Court.

"Politician makes a speech in 1980. It is recorded by a news channel. There is access. In 1992, he makes a different speech. A contradictory speech. Most people will have not have access to it. LLM can answer by prompt. Data cannot be copyrighted. It is impossible to grant injunction because the facts are not before you," he said.

Sibal said the technology is still evolving and that Court must not injunct the change.

"The copyright law could not have imagined large language models to be conceived in this fashion and being used for the purposes of change. Change is inherent in every aspect of life. You can't injunct change," he said.

The senior counsel contended that even if it is a commercial dispute, the decision should be in alignment with the constitutional rights.

"Please have Article 19(2) of the Constitution. None of the exceptions apply. I am entitled under Article 19 to protect my right to access information. It is my fundamental right," he added.

He said that restricting the same will amount to violation of free speech under Article 19.

"The public has right to access. An individual who sought that prompt has a fundamental right to find out about the contradiction. How will anyone research if someone has contradicted himself in a speech. If there is any law that restricts this right, it will be a violation of Article 19," he told the Court.

Sibal also argued that mere act of storage would not violate provisions of Copyright Act.

"The purpose of LLM is not to reproduce. The intent to store is to use the data. I must reproduce under Section 14 for copyright infringement. Act of storage is not infringement. What do I store? A speech, a math formula, data on climate and reproduce it, is it a copyright if I use it for progress of science?," he submitted.

In his arguments, Sibal also reflected on the news organizations using paywall to regulate access to news on their websites.

"How do you stop the circulation of ideas? If you say, whatever I have on my channel, you can't even look at, you can restrict it. It is all commercialisation, gone are the days of freedom of speech. I want to read an article from Indian Express, I can't. It is the only vehicle of political discourse so essential to democracy," he submitted.

Sibal said there was an element of public interest in promotion of LLMs as people are now able to swiftly access information.

"Earlier we had to go to libraries, now we can get information in a jiffy. It is so much easier now, this should be supported. You can't stall progress. At this stage, to grant any form of injunction will destroy the very process of change, which is inevitable."

The senior counsel also argued that the issue at hand involves a matter of public policy.

"As to how different jurisdictions in the global community use AI in the context of their peculiar interests and concerns is a matter of policy. Different jurisdictions may use it differently. Matter of policy is really in the domain of the government," he said.

At this stage, Justice Bansal asked,

"If the intent of the state is to encourage LLMs and use of artificial intelligence, what prevents the State from amending the copyright law?".

Sibal said that the government might do so.

Earlier, Senior Advocate Arvind Datar also argued for Broadband India Forum.

He submitted that any injunction would cause irreparable harm as LLMs enable the public, particularly in India, to have access to wider information. The large language model helps a lawyer, may help a doctor, a student appearing for a UPSC exam, he added.

"You can have transient storage. The LLM is accessing and integrating. The application for injunction should be dismissed. The Courts have held that in IPR matters, the nature of dispute may not be only between parties, it can have larger ramifications," Datar said.

The next date of hearing in the matter is November 21. The Court today indicated that another date for hearing would be in January 2026. The suit was filed in November 2024.

[Read live updates]