Meta Drops AI Training Mention From New Crawling Terms

Meta revised its forthcoming website-crawling terms late on Friday to emphasize its advertising and business products, dropping an explicit mention of AI training after criticism from business customers and media, according to Mercury News.

The company said the consolidated crawling terms applied only to its ads and business products and that this was clarified in the updated language. The word "train" no longer appears explicitly in the new terms, which focus on advertising and business products.

News/Media Alliance spokesperson Sam Quigley said the revised language limits what Meta may use publisher content for but does not limit its ability to crawl and collect that content. Quigley added that training is not expressly ruled out in the new terms and remains potentially possible.

Meta's statement said there would be no changes to how it uses web crawlers, but the new terms do not explicitly state this. The terms preview refers to data collection by automated software such as web scrapers, bots, robots, spiders, crawlers and user agents.

Background

Before the change, the policy stated that businesses using Meta products would be subject to terms letting Meta use their website data for essentially anything, including AI training. robots.txt has long been treated as a rule of the road for web crawler bots, and Google says agents get separate rules and that it will respect them.