OpenAI Refutes Claims of Using Indian Media Content for ChatGPT Training

While OpenAI's response suggests that it is compliant with existing regulations, the ChatGPT controversy has sparked a broader debate about media rights, data usage, and AI accountability.

OpenAI has denied allegations that it has used Indian media content to train its AI chatbot, ChatGPT. The company made this statement before the Delhi High Court, responding to copyright infringement claims filed by major Indian media houses. The case highlights growing concerns regarding the AI training process and the use of publicly available content for training large language models (LLMs). With Indian copyright laws evolving to address the challenges posed by artificial intelligence, OpenAI’s response to these allegations could set an important precedent for the ethical AI training debate.

OpenAI Denying Copyright Infringement

According to court filings reviewed by Reuters, OpenAI has formally rejected claims that it relies on proprietary Indian media content for training ChatGPT. In its statement, the company urged the Delhi High Court to dismiss the infringement claims brought forward by major media houses, including Mukesh Ambani-owned Network18 and Gautam Adani-owned NDTV.

OpenAI clarified that its AI training process is designed to comply with global ethical AI standards and does not involve the unauthorized scraping or reproduction of copyrighted content. The company also stated that it is under no legal obligation to enter into partnerships with media organizations to use publicly available information.

Also Read: OpenAI CEO Sam Altman to visit India amid legal and AI challenges

The legal dispute arises from concerns that AI chatbots, including ChatGPT, may be deriving insights and responses from copyrighted media content without authorization or compensation. Indian media companies have argued that AI models trained on their content could be monetized by tech companies, without proper attribution or licensing agreements.

This ChatGPT controversy is part of a larger global debate, as media houses worldwide demand regulatory clarity on how their content is used in AI training processes. Several organizations have begun exploring legal frameworks that compel AI companies to compensate content creators, a practice already seen in partnerships between news publishers and AI firms in certain jurisdictions.

This case raises important questions about Indian copyright laws in the age of AI. Currently, copyright frameworks in India do not explicitly address AI-generated content or the rights of content owners when their material is used for AI training. However, with increasing concerns from media companies, content creators, and regulatory bodies, there may soon be legal clarifications and policy updates.

For OpenAI, this case is significant in defining the scope of ethical AI training in India. If the court rules in favor of Indian media companies, it could set a precedent for AI firms to secure explicit licensing agreements before using news content for training. Conversely, if OpenAI’s argument prevails, it may reaffirm that AI models can train on publicly available content without requiring formal agreements.

As AI technology advances, discussions around ethical AI training and content rights will continue to evolve. Moving forward, there may be calls for Indian policymakers to establish clear guidelines on how Indian media content can be used in AI training processes. This could lead to the implementation of AI-specific copyright laws, requiring AI firms to compensate news organizations for using their data.

While OpenAI’s statement firmly denies any unauthorized use of Indian media content, the Delhi High Court’s ruling will be closely watched by AI developers, media houses, and policymakers alike. Whether this case leads to tighter AI regulations or broader content-sharing agreements, it is clear that the future of AI training and digital media rights is at a critical crossroads. 

Also Read: India to develop its own AI model like ChatGPT and DeepSeek in 2025