Meta has removed wording from its commercial terms that appeared to grant broad permission to use content from a business account’s websites and domains to develop, research, train and maintain its products. Legal experts had warned that the clause could cover publisher sites linked to Meta business accounts.
The language drew attention because publishers have challenged AI companies over the collection and reuse of copyrighted material. Although Meta prevailed in one prominent copyright case last year, other legal disputes involving AI training continue. The deleted provision therefore raised questions about whether accepting business terms could create a separate contractual route for scraping content beyond Meta’s platforms.
Meta says that interpretation was not its intent. A spokesperson said the update did not reflect any change in how the company crawls the web. In a later statement, Meta said the consolidated crawling terms applied only to its advertising and business products and that it had revised the document to make that scope clear. Removing the disputed wording resolves the immediate ambiguity but does not independently establish how Meta’s crawlers collect or use public web content. Publishers still need to evaluate the company’s technical behavior and the remaining terms rather than treating the revision as proof that no material is used for AI systems.