New York Times seeks sanctions against OpenAI

The New York Times and a group of publishers asked a federal judge in Manhattan on Thursday, July 9, to sanction OpenAI in their copyright case, accusing the company of misleading the court during discovery. The filing escalates one of the highest-profile lawsuits over how artificial-intelligence companies used news articles to build their systems. (nytimes.com) The publishers say the dispute is no longer only about whether AI training was lawful, but also about whether OpenAI was candid about what evidence it could search and produce. (msn.com) OpenAI has argued, like other AI companies, that training on digitized material found online can qualify as fair use under U.S. copyright law. ### What are the publishers accusing OpenAI of doing in discovery? (thehindu.com) The July 9 motion says OpenAI concealed its ability to search training datasets and model output logs while the plaintiffs were seeking evidence that their articles had been used or reproduced. Reuters reported that the newspapers accused OpenAI of lying to the court about its ability to search its systems for proof that it misused millions of articles in AI training. The New York Times reported that the publishers accused OpenAI of withholding evidence in the lawsuit. AP said the request for sanctions asked the judge to penalize OpenAI over what the outlets described as bad-faith conduct in discovery. (nytimes.com) ### Who is in the group pressing for sanctions? The New York Times is leading the group, and the New York Daily News is among the named participants. AP reported that other news organizations have joined the case over time, including MediaNews Group papers such as the Chicago Tribune, digital publisher Ziff Davis and the nonprofit Center for Investigative Reporting. (money.usnews.com) Reuters described the dispute as a high-stakes copyright fight in federal court in Manhattan. The case has become a focal point because it combines claims about AI training data with claims that OpenAI-generated outputs can reproduce publishers’ work. (nytimes.com) ### What does OpenAI say about using articles and books to train AI? OpenAI and other technology companies have argued that training AI systems on digitized books, articles and other writings found online is protected by fair use. AP said that defense is now being tested in dozens of lawsuits over training data. (apnews.com) That fair-use argument sits at the center of a widening set of cases brought by publishers, authors, artists and other rightsholders. In this case, though, the immediate dispute before the judge is procedural: whether OpenAI met its obligations to preserve, search and turn over evidence during discovery. (money.usnews.com) ### Why does the fight now reach beyond copyright claims? France’s competition authority ordered Meta Platforms on July 8 to resume talks with French media groups over payments for publishing content and to put forward a payment plan. Reuters reported that the authority said Meta likely abused its dominant position after publishers complained following the collapse of earlier negotiations. (apnews.com) The French order shows that publishers are also pursuing competition and payment disputes, not only copyright claims tied to AI training. The case concerns a year of alleged unpaid fees for the use of media content online, according to Reuters’ report on the regulator’s decision. (money.usnews.com) ### What happens next in the OpenAI case? The sanctions request now goes to the federal judge overseeing the Manhattan copyright case, who will decide whether OpenAI’s conduct warrants penalties. Reuters said the publishers asked for sanctions in the U.S. copyright dispute on July 9, while AP reported the motion seeks judicial action in the ongoing case. (thehindu.com) In France, Meta must present details of its payment plan within 15 days and resume negotiations with media groups, according to the competition authority’s order reported by Reuters. In New York, the next step is the court’s response to the publishers’ sanctions motion and any reply OpenAI files. (thehindu.com) (money.usnews.com)