ANI warns OpenAI's AI training on its content could harm news syndication

The Delhi High Court is hearing a lawsuit filed by news agency ANI, which accuses OpenAI of copyright infringement for allegedly using its content to train AI models like ChatGPT.

ANI has asked the court to prevent OpenAI from using its news content for AI training, warning that such practices could harm the news syndication industry. The agency also claims that OpenAI admitted to collecting its copyrighted content until October 2024.

An amicus curiae argued that AI is essential for innovation and questioned whether developing large language models (LLMs) without using copyrighted material is feasible. He also pointed to exemptions in the Copyright Act that might apply, suggesting that extracting facts from protected content does not necessarily violate copyright laws.

The news agency, however, argues that the company’s use of its content gives ChatGPT an unfair advantage, especially since the AI company has signed commercial agreements with global news organisations. The agency also accuses OpenAI of reproducing parts of its work verbatim or in similar form when responding to user queries.

The AI company has denied any wrongdoing, claiming that it makes fair use of publicly available data. It also challenged the Indian court’s jurisdiction, arguing that it has no physical presence in India and that its servers are located overseas. Additionally, it stated that U.S. laws require it to retain training data due to ongoing legal proceedings, which could conflict with any deletion order from the Indian court.

The Delhi High Court is scheduled to continue hearing the case on March 20.