In a significant development in the ongoing legal battle between AI company OpenAI and news agency ANI, OpenAI on Tuesday rejected allegations that it violated its previous undertaking regarding the use of ANI’s content.
The company asserted before the Delhi High Court that using ANI’s content for search purposes is different from using it for training its ChatGPT software.
The dispute centers around ANI’s copyright infringement lawsuit against OpenAI, which claims the AI company is using ANI’s content without proper licensing to train its large language model ChatGPT. During the hearing before Justice Amit Bansal, OpenAI’s senior advocate Amit Sibal argued that the company had not violated any court orders or copyright laws.
“The statement says that the content is blocklisted for the purpose of training. When it is for search, it is different from training,” Sibal told the court. He further maintained that OpenAI’s use of publicly available data for “non-expressive purposes” does not constitute infringement, arguing that “there cannot be a copyright in the news freely available.”
Indian News Agency ANI Sues OpenAI Over Content Scraping
This response came after ANI’s representative, advocate Siddhant Kumar, urged the court to rule in favor of the news agency. Kumar claimed that despite OpenAI’s previous undertaking to blocklist ANI’s domain, the company continues to scrape content from ANI’s subscriber websites to train ChatGPT and generate responses.
Kumar emphasized that licensing content to subscribers does not divest ANI of its copyright protections. “Merely if I have licensed content, I don’t cease to have copyright over those and my copyright is not divested,” he argued. He further noted that OpenAI had previously committed to blocking ANI’s domain and not using its content for future training of ChatGPT.
The issue has created widespread interest in India’s publishing and media sector. A host of organizations, including the Indian Music Industry, the Federation of Indian Publishers, and the Digital News Publishers Association, have come out in favor of ANI’s stand.
ANI Sues OpenAI Over Copyright
ANI, in its lawsuit, wants OpenAI to be prohibited from utilizing its content and to destroy all content it has accumulated. ANI contends that OpenAI has trained its language model on its content without securing a valid license, thereby commercially benefiting from its work.
While OpenAI has inked content agreements with international publishers like News Corp and The Guardian, which allow it to use their content for training as well as in responses by ChatGPT, the company has not inked such agreements with any Indian publisher.
Kumar also contested OpenAI’s “fair use” argument during the hearing on Tuesday, asserting the AI company cannot utilize ANI’s material for “commercial purposes” while benefiting from Section 52 of the Copyright Act’s protection. Section 52 normally permits fair use of copyrighted work for purposes of criticism, comment, news reporting, teaching, or research.
The decision in the case is likely to have wide-ranging implications for copyright law on AI-generated content and copyright protection of the original work of news agencies on the web. It would also set significant precedents on content licensing and content use rights across the fast-evolving area of artificial intelligence.
In November, while ordering OpenAI to accept the summons, the Delhi High Court declined to order the company to desist from using ANI content on an urgent basis after OpenAI told the court that it had already blacklisted ANI’s domain. The court also appointed legal experts as amici curiae, recognizing the intricate legal issues raised by new technological advancements.
The following hearing of the closely monitored case is scheduled for March 28.
