The AI Copyright Debate: 'Compensate, Don't Infringe'

On April 11, Times Now held the first of several discussions planned in the coming weeks in its endeavour to generate awareness about how the rights of content creators and news publishers are being infringed by AI applications. This violation of copyright norms is not confined to India. In Europe and the United States, media organisations have filed multiple lawsuits against companies, including Open AI and Alphabet Inc, for alleged copyright violations. Huge volumes of data and information - the original works of content creators who have a copyright over them - are used by AI applications to train large language models (LLMs) to provide responses to queries and prompts. According to Tollbit, a content licensing platform, news sites and blogs

receive 96 percent less traffic referrals from AI search engines than the traditional Google search. A research paper by Alex Webb and Amrutha Nair, titled The News Pipeline: Economic and Strategic Implications for AI for News Distribution, 99.7% of readers do not visit the original site when AI summarizes the news. According to Raptive, a publishing platform, the annual revenue loss as a consequence is estimated to be to the tune of US$2 billion annually for publishers.In India, the Copyright Act of 1957, protects the rights of content creators and publishers. Stakeholders from the media in India have also been engaging with the Indian government to ensure that AI applications, and the companies that own them, comply with Indian laws. News publishers, globally, have underlined the fact that copyrighted content used by AI platforms should be purchased, and that permission for use and payment of compensation are very much tied into the principle of copyright. In 2025, Anthropic arrived at a settlement valued at US$1.5 billion dollars following a lawsuit by several authors. In another development, the New York Times and Amazon signed a licensing agremeent for content to be used on Amazon's AI platforms. News Corp and Open AI had entered into an agreement in 2024. Is this the way forward?Moderated by Siddhartha Talya, the 'AI Copyright Debate' featured an engaging discussion on the subject.Pavan Duggal, Advocate and a Cyber-Law expert, said that the "fair use" defence used by AI companies to claim that they are not infringing copyright may not stand, and the provisions of the copyright laws in India are quite clear. He said: "Fair use and fair dealing are well-defined concepts. You cannot land up doing something which actually benefits you from a commercial standpoint, and yet argue that this is part of fair use or fair dealing. If it is for genuine criticism, analysis, examination purposes, commenting purposes, academic research, no problem. But the moment you use the copyrighted work of a copyright owner for the purposes of generating monies, even if you put your own elements therein, nonetheless, it still is impacting prejudicially the generation of revenues by the copyright owner. And that being so, under the copyright law, you are required to compensate the copyright owner. Therefore, the argument that because AI is a new transformational technology, the existing law will not apply is a legal argument that may not stand."Nikhil Pahwa, Founder and Editor of Medianama and a Digital Rights Activist, argued strongly in favour of compensation. He said: "What these models also do is that they use a mechanism called retrieval augmentation, a RAG model that is used to improve their outputs with fresh and relevant information, which means that something goes on the web, it'll come to your site, it'll come to my site, take information live so that it replaces what we provide to our readers immediately. So, there is substitution that's taking place without compensation. What is happening is that the cost of production of this information is being externalized to me, to you, and to all other media companies, and the benefit is being taken by the model that is taking from us without paying us. That cannot be allowed to stand whatever the Copyright Act says."

https://www.youtube.com/watch?v=7JkX8jJ-bTw

Michael Owens, President and CEO of the US Global Center for Cyber Policy, said that the law needs to catch up with the rapid change that AI has wrought. He said: "The path forward is not trying to stop AI, because I don't think that's a winning solution, but it's about modernizing what we're doing around copyright. So innovation at scale can work alongside these systems that will still pay people who make this work possible...when platforms are able to keep the users' information and the publishers end up losing, the business model is broken. The laws around the world are just going to have to catch up to the pace in which AI has changed the game."Ankush Sabharwal, Founder and CEO of BharatGPT and CoRover, said there was a strong moral basis for compensation as well. He said: "If a book is written, it's publicly available for the user to consume but if you're using it for some other purpose and monetizing it, theoriginal copyright holder must be compensated...not while the technology or the model is being created, but when it is being influenced and used and they [the AI app] are making money, they should be giving it back to content creators. Morally, I'm sure they must be compensated."