U.S. News & World Report sues ten OpenAI entities, alleging training on and output of its rankings and articles

What was filed

The complaint is Document 1 in No. 1:25-cv-09912, U.S. District Court for the Southern District of New York, filed November 26, 2025. It runs 69 pages. Counsel for the plaintiff is Rothwell, Figg, Ernst & Manbeck, P.C. The plaintiff demands a jury trial. (Compl. pp. 1, 4 to 5, 68 to 69.)

The plaintiff is U.S. News & World Report, L.P., which the complaint describes as a limited partnership headquartered in the district. The defendants are OpenAI, Inc.; OpenAI LP; OpenAI GP, LLC; OpenAI, LLC; OpenAI OpCo, LLC; OpenAI Global, LLC; OAI Corporation, LLC; OpenAI Holdings, LLC; OpenAI Foundation; and OpenAI Group PBC, referred to collectively as OpenAI. (Compl. pp. 1, 7 to 10, paras. 15 to 26.)

What the complaint alleges

It alleges that the usnews.com terms of use allow only noncommercial, personal use and bar use of the materials to develop or train a machine learning or AI system, including retrieval-augmented generation; that each publication carries copyright management information as defined in 17 U.S.C. 1202(c); and that usnews.com robots.txt disallows OpenAI's GPTBot and ChatGPT-User crawlers. (Compl. paras. 5 to 6, 85 to 86.)

On training data, it alleges more than 20,000 entries of U.S. News content in OpenAI's WebText dataset, and that the travel, cars, and real estate subdomains of usnews.com account for about 3.5 million, 3.8 million, and 410,000 tokens in C4, a filtered subset of a 2019 Common Crawl snapshot. (Compl. paras. 58, 60.)

On outputs, it alleges that ChatGPT reproduces U.S. News rankings and articles and attributes inaccurate information to U.S. News. Its example is a ChatGPT list of the top five children's hospitals for cancer in 2024-25 based on U.S. News rankings, which the complaint says included two institutions that were not in the top five, omitted two that were, and misordered three. (Compl. paras. 9 to 10, 80.)

Counts and relief

Count I alleges direct copyright infringement under 17 U.S.C. 501, based on building training datasets, training, storing the models, and disseminating output, and alleges willfulness. Count II alleges contributory infringement in the alternative, if end users are treated as the direct infringers. Count III alleges removal of copyright management information under 17 U.S.C. 1202(b)(1). Counts IV to VI allege trademark infringement and counterfeiting under 15 U.S.C. 1114 and false designation of origin and dilution under 15 U.S.C. 1125. Count VII alleges dilution under N.Y. Gen. Bus. Law 360-l. (Compl. paras. 131 to 196.)

The prayer seeks statutory and compensatory damages, restitution, and disgorgement; a declaration that the use is unlawful; a permanent injunction; destruction under 17 U.S.C. 503(b) of all GPT or other LLM models and training sets that incorporate the plaintiff's content; and costs and attorneys' fees. (Compl. pp. 67 to 68.)