The Print Army Comes for A.I.

David Remnick and Nick Thompson
Two decades ago, when Thompson was serving as Remnick’s digital chief, they adapted The New Yorker to Internet 1.0—which, as Remnick told me, required dramatic “psychological, editorial, and almost physiological” changes on the part of the staff. Photo-Illustration: Puck; Photos: Craig Barritt/Getty Images for The New Yorker/Bruno de Carvalho/SOPA Images/LightRocket via Getty Images
Dylan Byers
February 15, 2025

On Thursday, while top editors at Condé Nast, Vox Media, The Atlantic, and elsewhere were eagerly awaiting news regarding the finalists for this year’s National Magazine Awards—still a to-do in the editorial department, despite the honor having zero impact on the business—their legal teams were preoccupied with a lawsuit filed that morning in the Southern District of New York. The suit, brought by the News/Media Alliance on behalf of about a dozen other publishers, including Politico and the Los Angeles Times, accused the Toronto-based artificial intelligence company Cohere of stealing their content, using it to train its models, and making much of it free to users who would have otherwise hit the news organizations’ paywalls. When Cohere isn’t repurposing full articles, the complaint alleges, it also violates publishers’ trademarks by attributing “inaccurate” work to the publishers, undermining the standards of their brands. (Cohere has said it “stands by its practices for responsibly training its enterprise A.I.”)