A coalition of major news organizations has urged a federal judge to impose sanctions on OpenAI in a high-stakes copyright dispute over the use of published articles to train artificial intelligence chatbots. The request, filed in U.S. District Court in Manhattan, escalates a legal battle that could set precedent for how AI companies access copyrighted content.
Court filing demands accountability
The news outlets, including The New York Times, argue that OpenAI violated copyright law by using their articles without permission to train its GPT models. In a joint filing, they asked the judge to hold OpenAI in contempt for failing to preserve relevant evidence and to issue sanctions, including monetary penalties and an order to cease using the protected material. The filing states that OpenAI's actions have caused "irreparable harm" to the publishers' businesses.
The New York Times originally sued OpenAI and Microsoft in December 2023, seeking to end the practice of using published material for chatbot training. The lawsuit alleges that OpenAI copied millions of articles to build its AI systems, infringing on copyright and threatening the viability of journalism. OpenAI has defended its practices as fair use, arguing that training AI on publicly available text is transformative and legal.
Broader implications for AI and media
The case is closely watched as it could define the boundaries of AI training data. According to a report by the Reuters Institute, more than 40% of news publishers have blocked OpenAI's web crawler from accessing their sites since the lawsuit was filed. The news outlets' latest motion accuses OpenAI of destroying or failing to preserve evidence, including internal communications about training data selection. They seek a court order requiring OpenAI to preserve all relevant records and to produce documents showing how it selected training sources.
"We believe that OpenAI has deliberately evaded its obligations to preserve evidence," said a spokesperson for the news coalition, speaking on condition of anonymity due to ongoing litigation. "This is a matter of fundamental fairness and the integrity of the judicial process."
OpenAI's response and legal strategy
OpenAI has denied any wrongdoing, stating that it respects copyright and has implemented measures to allow publishers to opt out of training. In a statement, an OpenAI spokesperson said: "We are committed to working with publishers to ensure mutual benefit and have engaged in licensing agreements with many news organizations. We believe the court will find that our use of publicly available text is fair use."
The company has also argued that the news outlets' claims are overbroad and that AI training does not replace the need for original journalism. However, the publishers contend that OpenAI's chatbots can reproduce substantial portions of their articles verbatim, undermining their subscription and advertising revenue.
The judge has not yet ruled on the request for sanctions. A hearing is scheduled for later this month. The outcome could influence ongoing negotiations between AI companies and publishers over licensing fees and data usage.
