What was filed
The complaint is Document 1 in No. 1:26-cv-08892, U.S. District Court for the Southern District of New York, filed October 8, 2026. It runs 79 pages, not counting Exhibit A (the plaintiffs' copyright registrations) and Exhibit B (full versions of the output examples). Counsel for the plaintiffs is Rothwell, Figg, Ernst & Manbeck, P.C. The plaintiffs demand a jury trial. (Compl. pp. 1 to 2, 33, 78 to 79.)
The plaintiffs are USA TODAY Co., Inc.; Gannett Satellite Information Network, LLC; Gannett GP Media, Inc.; The Courier-Journal, Inc.; Des Moines Register and Tribune Company; Detroit Free Press, Inc.; Detroit Newspaper Partnership, L.P.; CA Florida Holdings, LLC; Scripps NP Operating, LLC; CA North Carolina Holdings, Inc.; GateHouse Media Oklahoma Holdings, Inc.; Journal Sentinel Inc.; GateHouse Media Ohio Holdings II, Inc.; and Phoenix Newspapers, Inc. The complaint says they are all owned by USA TODAY Co., Inc., which it says was formerly Gannett Co., Inc. (Compl. pp. 1, 12.)
The defendants are OpenAI Foundation; OpenAI GP, LLC; OAI International, Inc.; OpenAI OpCo, LLC; OpenAI Global, LLC; OAI Corporation; and OpenAI Group PBC. The complaint refers to them collectively as OpenAI. (Compl. pp. 1 to 2, 10 to 12.)
What the complaint alleges
It states that the plaintiffs own copyrights in content from 19 publications: USA TODAY, The Tennessean, Indy Star, The Bergen Record, The Enquirer, Asbury Park Press, Democrat & Chronicle, The Knoxville News-Sentinel, Naples Daily News, The Oklahoman, Milwaukee Journal Sentinel, The Columbus Dispatch, The Arizona Republic, The Courier-Journal, The Des Moines Register, Detroit Free Press, The Detroit News, The Palm Beach Post, and Star News. It says each publication includes copyright management information as defined in 17 U.S.C. 1202(c), and that registered copyrights are attached as Exhibit A. (Compl. p. 2.)
It alleges that OpenAI's large language models were trained on copyrighted material scraped from the internet without authorization, including the plaintiffs' content; that models memorize training data and can output verbatim or near-verbatim copies; that OpenAI used programs that strip CMI; and that ChatGPT and related products also output copies or extensive summaries through retrieval-augmented generation. It cites illustrative GPT-5.6 search results summarizing articles from those publications and points to Exhibit B for full versions. It also alleges that USA TODAY Co. terms of service bar AI training and substitutive AI summaries, and that the publications block OpenAI crawlers in robots.txt. (Compl. pp. 2 to 5, 25 to 61, 68 to 70.)
On WebText, it alleges more than 160,000 entries from the plaintiffs' domains, including 83,266 from usatoday.com. On a filtered Common Crawl subset known as C4, it alleges over 122 million tokens from those domains. It further alleges that Microsoft gave OpenAI a copy of the Bing Index, including the plaintiffs' content, in an initiative codenamed "Project Taxi," and separately ran a crawler, "Project Mango," on OpenAI's behalf, for which OpenAI paid Microsoft. (Compl. pp. 19 to 20, 27 to 29.)
Count I alleges direct copyright infringement under 17 U.S.C. 501. Count II alleges that OpenAI Foundation, OpenAI GP, LLC, OAI Corporation, OpenAI Group PBC, and OpenAI Global, LLC are vicariously liable because they directed, controlled, and profited from infringement by OpenAI OpCo, LLC and OAI International, Inc. Count III alleges removal of copyright management information under the Digital Millennium Copyright Act, citing 17 U.S.C. 1202(b)(1). The prayer seeks judgment against each defendant, jointly and severally, for statutory damages, compensatory damages, restitution, and disgorgement; a declaration; a permanent injunction; destruction under 17 U.S.C. 503(b) of all GPT or other LLM models and training sets that incorporate the plaintiffs' content; and costs, expenses, and attorneys' fees. The body of the complaint states that the plaintiffs seek damages in excess of $250 million. (Compl. pp. 5, 74 to 78.)