Meta’s AI could still scrape news sites, legal experts say. While Meta updated its business terms to remove language suggesting unrestricted training of its AI models on news sites like The Verge, legal experts argue these tweaks fall short of addressing their concerns. James Grimmelmann, a digital and information law professor at Cornell Law School, tells Nieman Lab that the new terms don’t sufficiently limit the “associated with your products or services” clause, which permits Meta to scrape entire sites. Initially, Meta’s terms for business accounts appeared to grant the company permission to train its AI on news sites, stating that Meta “may use any … to provide and improve our services.” However, the experts contend that the revised language still leaves room for Meta to freely scrape news sites, potentially leading to continued content usage without adequate compensation or recognition for publishers. The updated terms, effective from October 30th, may offer some relief, but the experts suggest that Meta could still leverage the broader “associated with your products or services” provision to scrape entire news sites, raising questions about the extent of the changes. Despite the tweaks, the legal landscape remains uncertain, with experts predicting that Meta’s AI could continue to scrape news sites with minimal constraints, possibly impacting publishers’ ability to control content usage and monetize their digital assets.