Publishers push 'compensate, don't infringe'

Publishers and creators are pressing artificial intelligence companies to pay for content, not just argue about whether scraping it was legal. (timesnownews.com) Times Now said it held a discussion on April 11, 2026 focused on copyright complaints from news publishers and creators, and framed the fight as one playing out in India, Europe, and the United States. The outlet said media organizations in Europe and the United States have already filed multiple lawsuits against artificial intelligence companies over training and output use. (timesnownews.com, courtlistener.com) At the same time, some publishers are cutting deals instead of waiting for courts. The Associated Press said in July 2023 that OpenAI licensed part of its text archive, and OpenAI and News Corp announced a multi-year partnership on May 22, 2024. (ap.org, newscorp.com) OpenAI also announced a global partnership with Axel Springer in December 2023, with selected reporting from Politico, Business Insider, Bild, and Welt used in ChatGPT responses with attribution and links. Those agreements gave publishers one model for turning archives into licensing revenue instead of uncompensated input. (openai.com) Newsrooms are also changing their own production lines as artificial intelligence content becomes harder to spot after publication. Anthony Paul wrote in Kaieteur News on April 12 that newspapers now face an “AI shock” and need stronger editorial systems to preserve trust and distinguish verified reporting from synthetic material. (kaieteurnewsonline.com) The pressure is not limited to newspapers. Reuters Institute said in January 2026 that news leaders expected “increased demand for verification work,” and Digital Content Next wrote in February 2026 that publishers adopting artificial intelligence were emphasizing policy, staff training, and human editorial judgment. (reutersinstitute.politics.ox.ac.uk, digitalcontentnext.org) Book publishing has already had a concrete example of what that looks like. Hachette Book Group pulled Mia Ballard’s horror novel *Shy Girl* in March 2026 over concerns that generative artificial intelligence had been used in the text, according to TechCrunch and Locus. (techcrunch.com, locusmag.com) That episode pushed the problem past courtroom briefs and into acquisition and editing workflows. Once a publisher has to halt a release, review a manuscript, or add extra checks before publication, artificial intelligence stops being only a copyright dispute and becomes an operations cost. (techcrunch.com, kaieteurnewsonline.com) Artificial intelligence companies argue that training on large datasets can qualify as fair use and say partnerships with publishers can support a “healthy news ecosystem.” OpenAI said in a January 2026 post about The New York Times lawsuit that courts had affirmed fair use in related rulings, while continuing to fight the Times in court. (openai.com, courtlistener.com) The split is now visible across the industry: sue, license, or build new filters into the newsroom. What publishers are rejecting is the idea that their work should train or feed artificial intelligence systems for free. (timesnownews.com, ap.org, openai.com)