ANI accuses OpenAI of using its news content without permission to train AI models, leading to unfair competition. The court debates whether AI’s data collection violates copyright laws.
Thank you for reading this post, don't forget to subscribe!
NEW DELHI: ANI told the Delhi High Court that OpenAI is using its news content to train artificial intelligence (AI) models, which is harming ANI’s business and creating unfair competition.
ANI filed a case saying that OpenAI is violating its copyright by using its content without permission.
The case was heard by Justice Amit Bansal.
ANI argued that ChatGPT is trained using ANI’s content, which is then stored and made accessible to users. However, OpenAI initially responded by saying that it does not operate in India and its servers are also not located in the country. Because of this, it claimed that no legal action should be taken against it in India.
During the court hearing, Amicus Curiae Arul George Scaria joined through video conferencing, while Amicus Curiae Adarsh Ramanujan was physically present in court. Advocate Adarsh Ramanujan, representing OpenAI, argued that AI language models are not designed to provide absolute truth but rather work based on predictive algorithms.
He explained that the answers given by AI models change depending on when the question is asked, proving that these models rely on language patterns rather than actual facts.
The court wanted to understand whether the AI model works by predicting responses instead of having real knowledge. Advocate Ramanujan stated that OpenAI’s AI does not directly use ANI’s content. He explained that the legal question is whether the way data is collected before training affects copyright rules.
He also pointed out that ANI’s content is often protected behind a subscription paywall, but ANI’s subscribers, depending on their subscription level, can republish the content.
Regarding public access to ANI’s content, Advocate Ramanujan argued that just because something is available publicly does not mean copyright laws don’t apply. He said that OpenAI only used publicly available data but also admitted that collecting and storing such data could lead to reproduction.
He also discussed Text and Data Mining (TDM), explaining that in some countries, exceptions are allowed for AI training and research. However, in India, there is no clear legal exception for TDM, apart from what is covered under fair use. He compared OpenAI’s method to that of search engines, explaining that collecting data is not just for storage but also for processing and analysis.
A major point of discussion was OpenAI’s “opt-out” system, which allows content creators to request that their data not be used for training in the future. Advocate Ramanujan argued that this system does not solve past copyright violations. He said that once an AI model is trained on a dataset, the training cannot be reversed, meaning ANI’s content may have already been used.
Advocate Sidhanth Kumar, representing ANI, countered by saying that ANI has journalists worldwide who work hard to produce original content. He explained that ANI’s subscribers are allowed to republish its news without changing it or using short excerpts, but they cannot sublicense or resell the content. He asked the court to decide whether simply making content publicly accessible gives OpenAI the right to use it for AI training.
During the previous hearing, Amicus Curiae Arul George Scaria stated that
“Courts should adopt a two-step approach. Firstly, we should decide whether the concerned use falls under exceptions like private/personal use, criticism/review, or reporting of current events. Secondly, we should employ a fairness analysis”.
For Plaintiff: Advocate Sidhanth Kumar
For Respondent: Senior Advocate Akhil Sibal
Case Title:
ANI Media Pvt Ltd v OpenAI Inc (CS COMM 1028/2024)
Would You Like Assistance In Drafting A Legal Notice Or Complaint?
CLICK HERE
Click Here to Read Our Reports on CJI Sanjeev Khanna
FOLLOW US ON YOUTUBE FOR MORE LEGAL UPDATES
