A Delhi high court division bench on Tuesday declined to immediately restrain OpenAI from scraping news agency ANI’s content to train ChatGPT and generate responses.
“We will not like to pass any order ex-parte without hearing the other side,” a bench of Justices Avneesh Jhingan and Manmeet PS Arora told ANI’s lawyer.
The court refused to pass the direction while dealing with ANI’s petition against a single judge’s July 24 order. In that order, the court dismissed ANI’s petition to restrain OpenAI from using its content to train ChatGPT, holding that the company’s use of such material falls within the statutory exception of “fair dealing” under Indian copyright law.
The ANI told the court that, from September 2024 to July 24, OpenAI had undertaken that it would not scrape content from the agency’s website and urged the court to direct OpenAI to continue with the same arrangement.
The court issued notice in ANI’s appeal and fixed December 5 as the next date of hearing.
In its appeal before the division bench, the news agency argued that the injunction had been denied without adequately considering OpenAI’s alleged admission that it had scraped, stored and commercially exploited the agency’s work. It added that such acts infringed its exclusive rights under the Copyright Act, including the rights to store, reproduce, adapt and translate its original works.
The agency argued that, despite finding infringement, the single judge’s ruling had broadened the scope of the fair-dealing exception for private use and research to effectively permit commercial exploitation of its copyrighted work.
