DNPA raises copyright storm in Delhi HC, alleges OpenAI is ‘erasing’ news media

India’s Digital News Publishers Association (DNPA) appeared before the Delhi High Court on August 5 to argue that OpenAI is infringing the copyright rights of Indian media outlets through unauthorised use of news content in its AI models. The case reflects growing tension between global AI platforms and local journalism businesses.

During the hearing, DNPA lawyers stated that news publishers could vanish if ChatGPT continues using their content without consent or compensation. They accused OpenAI of scraping articles, storing them in models, and allowing end users to reproduce news without attribution or licence fees. "Print and digital news will vanish, only ChatGPT will remain," the DNPA warned.

The legal action comes on the heels of earlier litigation by Indian news agency ANI, which filed suit in November 2024 over similar allegations—that ChatGPT used its copyrighted reports without permission or licensing. DNPA represents digital arms of major media businesses including the Hindustan Times, Indian Express, Network18, NDTV and Dainik Bhaskar, among others.

OpenAI has responded with a February legal filing denying it used any content from DNPA members or ANI to train its models, asserting that it relies solely on publicly available data protected under fair-use principles. The company also contends the Delhi High Court lacks jurisdiction over its servers or operations based abroad.

The core of the dispute centers on whether Indian copyright law extends to generative AI training and whether data scraping without licensing constitutes infringement. DNPA argued that unscrupulous use of news content undermines the sustainability of journalism, pointing to declining revenues and weakened local media ecosystems.

For Indian marketers and media strategists, the case raises critical questions about content attribution, rights licensing and the future of AI-powered media workflows. A ruling in favour of DNPA could oblige OpenAI to seek licensing agreements in India—aligning local practices with its deals abroad with publishers like the Financial Times and Time magazine.

If courts find that generative AI needs licensing under existing copyright frameworks, it could trigger broader ripple effects across platforms like Google Bard, Meta’s LLaMA‑based services, and others using similar training methodologies. It could also help codify digital journalism’s economic value in the age of AI. The next hearing in the ANI‑OpenAI litigation is expected in January 2026.