400 newspapers file a lawsuit against OpenAI and Microsoft regarding AI.

      A coalition representing nearly 400 local newspapers in the US has filed a lawsuit against OpenAI and Microsoft, claiming that AI training on their journalism is detrimental to local news. This marks the largest copyright lawsuit the local press has initiated to date.

      Local newspapers provide coverage of meetings that algorithms do not attend, such as council votes, school board disputes, obituaries, and new downtown restaurants. The owners of these nearly 400 papers are seeking a court ruling to establish the value of their reporting to an AI company.

      On June 24, a nationwide coalition of publishers filed the lawsuit against OpenAI and Microsoft in a federal court in Manhattan. The 55-page complaint states that the two companies copied hundreds of thousands of articles to create ChatGPT and Microsoft Copilot without seeking permission or providing payment.

      Details of the case were initially reported by Courthouse News, with Richner Communications, a Long Island publisher, leading the charge. The publishers are represented by Matthew Platkin, who was New Jersey's attorney general from 2022 to 2026 and now operates his own firm. The lawsuit includes three counts of copyright infringement and seeks statutory damages, actual damages, a return of profits, and compensation for legal fees.

      The significance of this case lies in its scale—it represents nearly 400 newspaper titles in a single filing. More importantly, it showcases local and regional papers banding together in a way that has not been seen before. These are the underfunded publications that cover local courts and government.

      The plaintiffs listed in the lawsuit span a broad spectrum of regional newspapers, including the Arkansas Democrat-Gazette, the New York Amsterdam News, Newspapers of New England, the Ogden Newspapers chain, and Straus Newspapers, with their total number of titles reaching into the hundreds.

      The publishers assert that OpenAI and Microsoft “systematically and secretly crawled” numerous news sites, including those with paywalls, and copied the articles onto their servers. They also claim that these companies removed essential copyright management information such as bylines, publication names, usage terms, and copyright notices.

      This removal is central to a secondary claim, as the publishers argue that stripping this information functioned as a crucial part of the defendants’ data ingestion process, effectively disconnecting each article from its original source. They contend that the revised text subsequently trained large language models, which “memorized” and reproduced the articles verbatim in response to users.

      The publishers maintain that removing ownership data is illegal under the Digital Millennium Copyright Act. According to the New Jersey Globe, they also reference testimony from OpenAI chief Sam Altman to the British House of Lords, where he stated that training contemporary models would be “impossible” without using copyrighted material.

      Matthew Platkin emphasized the limitations of AI, stating, “AI systems do not critically evaluate city council and community meetings,” nor do they investigate local crimes, publish obituaries, or cover new restaurants. Local journalists perform these vital functions.

      Economic considerations amplify this argument. Local newspapers operate with limited budgets and dwindling staff. The complaint describes the unauthorized use of their content as a “death knell” for a sector recognized as one of the most trusted in the US. The assertion is that the harm caused by this copying affects the outlets that are least capable of managing it.

      OpenAI has faced similar challenges before, including a lawsuit from Encyclopaedia Britannica and its Merriam-Webster division regarding “massive copying” of their materials, alongside ongoing litigation with the New York Times. The newspapers' lawsuit is part of a broader movement among rights holders, including authors and actors, who are working to prevent AI from utilizing their work without consent. Some, like Getty Images, have opted for agreements rather than litigation.

      Microsoft is named in the lawsuit because it is deemed “an indispensable partner in virtually every aspect of OpenAI’s commercial enterprise.” This partnership began in 2019 with a $1 billion investment, fundamentally altering both companies. Neither organization commented on the lawsuit when it was filed.

      OpenAI has consistently maintained that training on publicly available text qualifies as fair use, a defense it is employing in all similar cases. The publishers believe that the covert nature of the data collection and the absence of bylines undermine that defense, aiming for a jury to view the act of crawling behind paywalls as theft rather than research.

      As OpenAI moves toward an $852 billion valuation, fueled by the belief that its models can provide answers to almost any query, this case poses a more focused inquiry: if those answers were derived from reporting that went unpaid, who bears the financial responsibility? A judge in Manhattan will resolve this issue, independent of any product strategy.