Starting October 30, Meta may scan company websites
Meta's new commercial terms will apply to every company that uses Facebook, Instagram, WhatsApp or Messenger for business purposes. The collected content may be used, among other things, to train the company's products.
Meta is updating its Commercial Terms, that is, the terms for business accounts. The change takes effect on October 30, 2026. Companies using Meta's products for commercial purposes will grant it permission to crawl, that is, to automatically browse and retrieve the contents of domains, websites and URLs associated with their products or services. The change was first described on October 16 by Nieman Lab, followed by, among others, Ars Technica and Search Engine Journal.
According to Search Engine Journal, until now this right applied only to advertisers who had placed an advertising order. The previous version of the terms from March 6, 2024 did not contain a paragraph on crawling. The new terms will cover all companies running pages on Facebook, using Business Suite (Meta's tool for managing business accounts) or ads, as well as publishers, regardless of ad spending. The clause also allows the creation of test login data. James Grimmelmann, a law professor at Cornell Law, believes this makes it possible to bypass registration requirements and paid content access barriers (paywalls).
The downloaded content may be used to develop, research and train (“train”) Meta's products. The word “AI” does not appear in the terms, but lawyers point out that this also concerns training artificial intelligence models. Matt Topic of the Loevy & Loevy law firm notes that access to such materials would probably cost many millions of dollars on the licensing market.
Meta denies this interpretation. Vice President of Communications Andy Stone wrote on X on October 16 that this is about retrieving information on products sold on external websites for ad and content recommendations, not about acquiring content from news services. Another spokesperson says it is about organizing existing provisions and reminds of the robots.txt file, in which a website owner tells bots what they are not allowed to retrieve. However, the new terms do not mention robots.txt, and lawyers warn that the agreement may take precedence over such instructions. According to Nieman Lab, Meta did not tell the researcher who discovered the clause where this provision had previously existed.
Meta does not provide an option to opt out of the clause itself. A company that does not accept the changes is expected to stop using Meta's products for business purposes. Continued use after October 30 means consent, which may also cover the company's website, not just its social media profile.