Despite the changes, concerns remain. TechTechPosts will be added to your daily feed and main feeds. Follow See All TechAIAIPosts for this topic.

Follow See All AINewsNewsPosts for this topic. Follow See All NewsMeta’s AI may still scrape websites, according to legal experts. Emma Roth, Emma RothNews WriterPosts will be added to your daily feed and main feeds.

Follow See All by Emma RothOct 22, 2026, 8:32 PM UTCIllustration by Nick Barclay / The VergeEmma RothEmma RothPosts will be added to your daily feed and main feeds. Follow See All by Emma Roth is … Before, it was a writer and editor at MUO. Meta has modified its negotiation conditions to remove language suggesting that Meta could use its AI models freely on websites like The Verge — but legal experts say these changes are not sufficient to resolve their concerns. ‘This does nothing to limit the “associated with your products or services” [language] that would allow Meta to scrape entire sites,” James Grimmelmann, professor of digital law and computer science at Cornell Law School, tells Nieman Lab.

Sometimes, the new Meta negotiation conditions appeared to authorize Meta to use its own AI models on websites that were messy. The conditions, which entered into force on October 30, were highlighted that Meta “could use any content” originating from websites associated with a negotiating party for “development, research” and “train” its products. Legal experts told Nieman Lab that this could allow Meta “to argue that its negotiation conditions exceed other blocking instructions, such as the robots.txt file that has always said what can be scraped (like Google) what can be scraped).

Subsequently, Meta has removed the language suggesting that Meta would “develop” its products with scrapable content. However, the modifications may not be sufficient to prevent the possibility remaining. ‘The updated version of the new negotiation conditions does not seem to need to prevent Meta using that scrapable content to develop any ‘business product.’ This is not a clearly defined term, and presumably Meta is developing business models AI.” I thought ‘develop’ was broad enough to include ‘train,’” Matt Topic, media and intellectual property owner at Loevy & Loevy, tells Nieman Lab. Meta has maintained that the new negotiation conditions are not designed to scrape news content for free.