Guides

Meta's 30 October Terms and the Crawl Clause TechCrunch Says Was Removed

18 October 2026 · RenderBob team

Meta's Commercial Terms take effect on 30 October. A clause let it crawl business-linked sites for training. TechCrunch now says that clause was removed. Read the live text.

Meta's revised Commercial Terms take effect on 30 October. Nieman Lab and Search Engine Journal described a new paragraph that let Meta crawl websites tied to a business account and use the content to train its products. On 17 October, TechCrunch reported that The Verge says that provision has been removed. We have not read the live terms.

What the clause said

Search Engine Journal, drawing on the draft terms and on Nieman Lab, describes permission for Meta, or anyone acting for it, to crawl domains associated with a business's products or services, and URLs made available through that business's activity on Meta's products. The listed uses include developing, improving, researching, and training those products. The clause also covers creating test credentials for those domains. Continued use counts as agreement. The alternative written into the terms is to stop using Meta's products for business, which covers Pages, ads, and the business dashboard. The Commercial Terms dated 6 March 2024 had no crawling clause. Self-serve ad terms already allowed Meta to crawl advertiser sites for uses that include research and training.

What Meta and the lawyers said

Nieman Lab said Meta declined to answer its questions. Andy Stone called reports that the terms allow scraping of news sites false, and said the point is accurate product information for ad and content recommendations. TechCrunch later carried Meta's statement that the crawling language they consolidated applied only to ads and business products, that the terms have been clarified, and that this does not change how Meta's crawlers work. A spokesperson told The Indian Express that owners can still use robots.txt, and that the update does not grant rights that override those preferences.

Nieman Lab quoted three lawyers. Matt Topic said it is unclear a court would enforce the provision, and that publishers should assume it would. James Grimmelmann said the clause could be read as permission to train on publishers' content, and that test credentials could get past a registration wall. Benjamin Sobel compared the structure to tying in antitrust. Whether training is fair use is a separate question, and it is not settled.

Removed, or only rewritten

TechCrunch, citing The Verge, says the provision was removed after the backlash. PPC Land's article of 16 October, written before that report, still had the clause in. Read the terms on the page you would be accepting. A news summary is not the sentence that binds the account.

What to check if you run business accounts

• List the domains connected to your Pages and ad accounts.
• Keep unreleased material, including screeners, behind a login. Do not rely on robots.txt alone.
• Decide whether the press kit, the portfolio, and the reference pages should be crawlable at all.
• The effective date is still 30 October. Open the live terms that day. TechCrunch says the broad clause came out. Confirm the sentence yourself.
• Ask counsel whether a commercial contract can override a robots.txt instruction. The ad terms already mention research and training.

What to watch

The text that is actually live on 30 October, any formal opt-out, and whether other platforms add the same kind of clause.