Eight U.S. newspapers sued Microsoft and OpenAI in federal court on April 30, 2024, alleging that the companies used millions of articles without permission to develop and operate AI products including ChatGPT and Microsoft Copilot. The complaint also cited alleged outputs that reproduced article text or falsely attributed material to the papers. Those claims are allegations, not findings that either company infringed copyright.
Which newspapers filed the lawsuit?
The case, Daily News LP et al. v. Microsoft Corporation et al., was filed in the U.S. District Court for the Southern District of New York as No. 1:24-cv-03285. Daily News, LP, publisher of the New York Daily News, was the lead plaintiff. The eight plaintiffs were:
- New York Daily News
- Chicago Tribune
- Orlando Sentinel
- South Florida Sun Sentinel
- San Jose Mercury News
- Denver Post
- Orange County Register
- St. Paul Pioneer-Press
Contemporary coverage described the papers as part of newspaper operations controlled by Alden Global Capital, including MediaNews Group and Tribune Publishing. The filed complaint names Microsoft and multiple OpenAI entities as defendants; the case docket identifies Judge Sidney H. Stein.
What did the publishers allege?
The newspapers alleged that Microsoft and OpenAI copied millions of copyrighted articles without permission or payment, used them in developing and operating AI systems, and commercially benefited from that work. The complaint discussed ChatGPT and Copilot and alleged that the systems could produce lengthy passages or material resembling specific articles. It also raised concerns about false or fabricated material attributed to the publications.
#1 Best Overall
The publishers’ broader argument was that their reporting represents a substantial investment, and that AI products could draw on or reproduce that work while competing with the publications that created it. The complaint is not proof that every plaintiff article appeared in a particular training dataset, or that every cited output infringed copyright.
Why the publishers say it matters to their businesses
The alleged harm is not limited to whether a chatbot repeats a passage. If an AI product gives readers useful answers without sending them to a news site, publishers argue, it could reduce visits that support advertising and subscriptions. Unlicensed use could also affect the value of content licensing and publishers’ control over reuse of their archives. These are the plaintiffs’ asserted commercial concerns, not established findings about the financial effect of these products.
Rank #2
- Ideal for Gifting
- Ideal for a bookworm
- Compact for travelling
The dispute therefore concerns who may capture value from reporting: the organizations that pay to gather and publish news, or the AI companies that use news content to build commercial services. How much any particular AI response substitutes for a publisher visit, and how that affects a publisher’s market, would require evidence rather than assumption.
Training, web retrieval and copied outputs are different issues
The complaint and related coverage touch on several technical and legal questions that should not be treated as interchangeable:
- Training-data copying: The allegation that articles were copied and used in developing models. Whether a work was copied, how it was acquired, and whether a particular use is lawful are separate questions.
- Live retrieval: A system may fetch or reference material from the web while answering. The newspapers alleged that Copilot could retrieve recent pages and reproduce article text; that is distinct from a model generating text from learned patterns.
- Memorization or regurgitation: A model may produce text resembling or matching material it has encountered. An example of an output does not by itself establish that the specific article was in a training dataset or determine whether the output is infringing.
- Copyright in an output: Reproducing protected expressive passages can raise a different question from whether copying for training was permitted.
- False attribution: A system may invent or misattribute information. That can create accuracy and reputational concerns, but a fabricated attribution alone does not prove copyright infringement.
Attribution and permission are also distinct: naming a newspaper does not itself grant a license. And factual information generally raises different copyright questions from a reporter’s original expression, such as a distinctive passage or article structure.
How OpenAI and Microsoft responded
In its initial response reported by the Associated Press, OpenAI said it supported news organizations and was engaged in partnerships and discussions with publishers. Microsoft declined to comment to the AP at the time. In related litigation, AI companies have argued that using publicly accessible material to train models may be fair use under U.S. copyright law. OpenAI has also argued in related proceedings that some plaintiffs engineered prompts to elicit memorized text; that is the company’s position, not a court finding about the examples in this case.
Rank #4
Fair use is a defense assessed under the circumstances, including the purpose and character of the use, the nature of the copyrighted work, the amount used, and the effect on the work’s market. The claim that material was publicly viewable does not, by itself, settle whether a particular commercial use was authorized or fair. The application of copyright law to AI training and outputs remains contested.
How this case relates to the New York Times lawsuit
The eight-newspaper action is separate from The New York Times lawsuit against OpenAI and Microsoft, filed in late 2023. The cases overlap in their focus on alleged use of journalism to train AI systems, and the eight-paper docket identifies the matter as related to the Times litigation. But they have different plaintiffs, pleadings, examples, and procedural histories; they should not be described as one lawsuit.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsBest Value
- It can be a gift option
- Comes with secure packaging
- Helpful in various ways
What is the case’s status?
The docket records motions to dismiss, opposition filings, and a discovery-planning report by June 28, 2024. Later, Judge Stein allowed the bulk of the broader newspaper copyright litigation involving OpenAI and Microsoft to proceed while dismissing some claims, according to the Associated Press report on the ruling. Allowing claims to proceed is not a finding that infringement occurred, and the information available here does not establish a final liability judgment in this eight-newspaper case.
What publisher licensing deals do—and do not—settle
OpenAI has entered licensing or partnership arrangements with some publishers, including the Associated Press, Axel Springer, Prisa Media, Le Monde, and the Financial Times, as reported by the Associated Press. A license is contractual permission, and its scope depends on the agreement: it may treat archives, current articles, search display, training, or product development differently.
An agreement with one publisher does not establish permission from another. Nor does a licensing deal automatically resolve a claim about earlier use; whether it affects a particular claim depends on the agreement and the legal circumstances. Licensing is one way to negotiate access, while fair use is a separate statutory defense.
What the dispute could mean for AI and news
The case puts several questions with practical consequences before the courts: whether and under what conditions copyrighted news can be used to train commercial AI systems; when outputs cross the line into reproducing protected expression; and how live retrieval, attribution, and safeguards affect the analysis. Its outcome could influence licensing negotiations, dataset practices, chatbot citations, and efforts to prevent outputs from substituting for publisher content.
For newspapers, the stakes include traffic, subscriptions, advertising, and the bargaining value of reporting archives. For AI companies, the central issue is whether large-scale training can proceed without individual licenses under a fair-use theory. The court’s eventual treatment of the facts and claims—not the existence of the complaint alone—will determine how much this case answers those questions.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
