Meta Drops Training From Its Crawl Clause And Keeps The Crawl

Meta has revised its Commercial Terms to remove explicit permission for using crawled content in AI model training. The updated clause restricts the stated purposes of web crawling to improving advertising tools, business products, and maintaining platform security. This change directly addresses concerns from publishers and legal experts who feared the original wording granted broad rights to scrape news sites for artificial intelligence development. Despite this adjustment, Meta retains its general authority to crawl websites associated with business products. Legal analysts remain skeptical, arguing that terms like "develop" are sufficiently broad to still encompass training activities. Furthermore, the revision does not explicitly reference robots.txt signals, leaving ambiguity regarding whether businesses can effectively opt-out of specific types of data collection, such as AI training, through standard technical controls. This situation is critical for open data advocates and publishers because it highlights the tension between corporate data access rights and publisher control. The ambiguity suggests that relying solely on robots.txt may not be legally enforceable enough to prevent unauthorized data scraping. Companies must carefully review their technical settings and understand that acceptance of these terms may still imply significant data usage rights for Meta, impacting how open data flows from publisher sites to tech platforms.

Source: searchenginejournal.com
Published on 2026-10-20

Meta Drops Training From Its Crawl Clause And Keeps The Crawl

Related news