Policy & Regulation
Meta drops AI-training clause from business terms
Meta has removed a commercial-terms clause that legal experts said could let it scrape partner websites to train its AI models, after reports flagged the wording.
Meta recently rewrote part of its commercial terms of use to say it could “use any content from [a business account’s] domains, web pages, or URLs to provide, develop, improve, research, train, and maintain the integrity of” its products and services. Legal experts told Nieman Lab the wording looked broad enough to let Meta scrape any publisher’s site tied to a Meta business account and feed that content into its AI training — even as a Meta spokesperson insisted the update didn’t actually change how the company crawls the web.
Now, following reporting from The Verge, Meta has quietly dropped the clause. The company says the language it had folded in was meant only for its ads and business products, and that its actual web-crawling practices haven’t changed.
Why it matters
The reversal shows how closely publishers and AI companies are now scrutinizing the fine print governing whether content can be scraped for AI training — and how quickly a provision can get walked back once it draws outside scrutiny.