OpenAI faces sanction request from publishers

The New York Times, the New York Daily News and other publishers asked a federal judge in Manhattan on Thursday to sanction OpenAI in their copyright case, escalating a fight that has moved beyond whether AI training used protected journalism and into what evidence OpenAI preserved and disclosed. The publishers said OpenAI misled the court about its ability to search training datasets and ChatGPT output logs for copies of their articles. They also said the company deleted or rendered unsearchable billions of relevant ChatGPT conversations. OpenAI did not immediately respond to requests for comment, according to Reuters and other reports. ### What are the publishers accusing OpenAI of hiding? The July 9 sanctions motion says OpenAI told the court it could not search its large language models, datasets or output logs for the publishers’ copyrighted material while concealing that it had already run such searches, Reuters reported. The publishers argued that discrepancy goes to the center of discovery because the case turns in part on whether OpenAI copied and retained news content during model training and output generation. (money.usnews.com) Ian Crosby, lead counsel for The New York Times, said in a statement that “for over two years” OpenAI had lied to the Times, the Daily News plaintiffs, the public and the court. Crosby said OpenAI had claimed searches for copied news content in ChatGPT outputs were infeasible, burdensome and invasive of user privacy while concealing that it had already done them. (money.usnews.com) ### What sanctions are they asking the judge to impose? The publishers asked for sanctions that include attorneys’ fees and a court finding that OpenAI’s chat logs showed misuse of their copyrighted works, Reuters reported. That request matters procedurally because sanctions can go beyond a reprimand and affect how facts are treated later in the case. (money.usnews.com) The Associated Press reported that the newspapers framed the issue as one of withheld training-data evidence and noncompliance with discovery obligations. ABC, carrying the AP report, said the dispute could shape the future relationship between AI companies and publishers because it focuses on what records an AI developer can produce about how its systems were built. (money.usnews.com) ### Which case is this, exactly? The underlying litigation began when The New York Times sued OpenAI and Microsoft in 2023, accusing them of using millions of Times articles without permission to train the model behind ChatGPT, Reuters reported. A separate publisher suit led by the Daily News was filed on April 30, 2024, in the Southern District of New York against Microsoft and multiple OpenAI entities. (apnews.com) In April 2025, the Judicial Panel on Multidistrict Litigation transferred related actions from New York and Northern California to the Southern District of New York for coordinated pretrial proceedings. CourtListener’s docket shows the consolidated proceeding is assigned to U.S. District Judge Sidney H. Stein, with Magistrate Judge Ona T. Wang designated for referred matters. (money.usnews.com) ### Why does the fight now center on logs, datasets and search tools? The publishers’ motion, as described by Reuters, targets OpenAI’s statements about whether it had tools to search datasets and output logs for copyrighted material. Reuters said an OpenAI employee later testified that the company had “performed multiple searches” for the news plaintiffs’ content, a point the publishers used to argue that OpenAI’s earlier representations were false. (courtlistener.com) OpenAI has separately defended itself in public filings and on its website by arguing that AI training is fair use and by criticizing demands for large volumes of ChatGPT user conversations on privacy grounds. That public position is distinct from the publishers’ new claim that the company failed to preserve or disclose relevant evidence in discovery. ### What happens next in court? (money.usnews.com) The case remains in pretrial proceedings in the Southern District of New York, where Judge Stein is overseeing the consolidated litigation and Wang handles referred discovery matters, according to the docket. The next step on this issue is the court’s consideration of the sanctions motion and any response OpenAI files to contest the publishers’ account. (courtlistener.com) (openai.com)