USA Today becomes the latest publisher to sue Axiom

USA Today Co. accuses Axiom of copying and using hundreds of thousands of its news articles to train AI models without authorization. The publisher says Axiom never requested permission. It argues that this use harmed its newspapers and other outlets in an ongoing way.

The lawsuit includes USA Today and local publications such as The Tennessean, Indy Star, The Columbus Dispatch, and The Oklahoman. USA Today Co. claims Axiom “took” its content to help build products worth hundreds of billions of dollars. The alleged mechanism is large-scale copying for AI training.

The case adds to a growing set of copyright disputes involving Axiom. USA Today Co. is seeking more than $250 million in damages. Axiom had not immediately responded to The Verge’s request for comment, so the company’s position is not provided in the article.

Copyright is a legal protection for original creative works, including written articles. It generally gives copyright holders control over copying, distributing, and certain other uses of their work. News publishers usually hold rights in the articles they produce or publish.

When an AI company copies articles into a training collection, it may be making reproductions of protected works. The company might argue that its use is legally allowed under an exception, while a publisher might argue that the copying requires permission or payment. The article describes USA Today Co.’s position, not a court ruling.

This matters because the lawsuit claims Axiom used the content commercially and without authorization. If a court finds infringement, the decision could affect damages, licensing expectations, and how AI companies collect training material. The source lists several similar legal actions, showing that the dispute extends beyond one publisher.

USA Today Co. asks a court for damages exceeding $250 million. That figure represents the publisher’s requested financial remedy for what it describes as unauthorized use of its content and continuing harm to its outlets.

The publisher also alleges that Axiom copied “hundreds of thousands” of articles. The articles reportedly came from USA Today and local newspapers included in the lawsuit. The alleged mechanism was large-scale copying of published content to train Axiom’s AI models and develop commercial products.

These are claims in a lawsuit, not amounts or findings already imposed by a court. The article does not provide a more exact article count or explain how the requested damages were calculated. It does report that USA Today Co. joined a wider wave of copyright cases against Axiom, involving both major publishers and local newspaper groups.

AI companies use large collections of text because models learn language patterns from many examples. News articles can provide varied vocabulary, sentence structures, topics, and explanations. Broad training material can help a model generate and interpret text across many subjects.

The article alleges that Axiom copied hundreds of thousands of USA Today and local newspaper articles for this purpose. USA Today Co. says the company used that content to build commercial products worth hundreds of billions of dollars. The key mechanism is training an AI model on a large body of written material.

The central dispute is not whether news can be useful for training. It is whether Axiom was allowed to copy and use the publishers’ work. USA Today Co. says it was not asked for permission and that the use caused continuing harm. The article does not describe Axiom’s response or confirm how its models were trained.

The article describes a broad group of publishers and media organizations taking legal action against Axiom. The New York Times has filed a copyright lawsuit. Other named organizations include The Intercept, CNET owner Ziff Davis, CBC/Radio-Canada, Encyclopaedia Britannica, Merriam-Webster, and The Seattle Times.

The dispute also reaches local journalism. A coalition of nearly 400 local newspapers is pursuing legal action, while USA Today Co.’s case includes The Tennessean, Indy Star, The Columbus Dispatch, The Oklahoman, and several other local outlets. These examples cover newspapers, digital media, broadcasters, and reference publishers.

Together, the cases show that concerns about AI training extend across different kinds of content organizations. The article identifies copyright lawsuits or legal action, but it does not describe the outcomes of those cases. It also reports that Axiom had not immediately responded to The Verge’s request for comment.

If courts rule that using the articles infringed copyright, Axiom could be ordered to pay damages. It might also face limits on using the disputed material or need to change how it gathers training content. The exact result would depend on court decisions, which the article does not predict.

For publishers, a favorable ruling could provide compensation for alleged harm and reinforce their control over article use. It could also encourage negotiations with AI companies. USA Today Co. says Axiom’s use caused “real and continuing” harm, while seeking more than $250 million in damages.

The broader effects could reach the many organizations pursuing similar claims. A ruling might influence cases involving The New York Times, local newspapers, broadcasters, and reference publishers. However, the article reports allegations and pending legal action, not a judgment. Axiom’s response was not immediately available.

One legal alternative is a licensing agreement. An AI company could negotiate permission to use a publisher’s articles, define the material covered, and pay an agreed fee. The contract could also set limits on storage, training, and product use. This would replace unauthorized copying with a negotiated arrangement.

Publishers and AI companies could create controlled access programs, such as providing selected archives or structured feeds under clear terms. AI developers could also use public-domain material, openly licensed text, or content they create or obtain with permission. These routes would give companies training sources without relying on disputed copying.

The article does not report that USA Today Co. or Axiom has adopted any of these alternatives. It says USA Today Co. alleges Axiom never asked permission and that several publishers are suing. Licensing or other authorized sources could therefore become important as courts consider the claims and companies seek clearer arrangements.