Meta's new commercial terms take effect October 30: it may crawl websites linked to business accounts to train AI, with no opt-out

Start with the clause itself and its effective date. Nieman Lab reported it first on October 15, with The Verge and Ars Technica following on October 16: Meta's updated commercial terms take effect on October 30. The text says business accounts authorise Meta, or any entity acting on its behalf, to use automated software such as web scrapers, bots, robots, spiders, crawlers and user agents and to create test credentials in order to access, index, cache, analyze, collect, retrieve or crawl the domains, web pages or URLs associated with their products or services, or made available through their or users' use of Meta company products and services, and to periodically pull updates from them. Meta may use any content from those domains, web pages or URLs to provide, develop, improve, research, train and maintain the integrity of Meta company products, services and features. The scope covers business accounts that promote their content across Facebook, Instagram, WhatsApp and Messenger, which Nieman Lab notes means virtually every news organisation has at least one, Nieman Lab included.

The second point is Meta's own framing. In a blog post, Meta says the update streamlines its commercial terms, adding that the changes reduce overlap between its agreements and improve the clarity of the terms that apply to your use of its products. A public affairs spokesperson told Nieman Lab by email that language around crawling had previously been scattered across different terms of use, including terms about using generative AI in ads, and that this update consolidates them in one place; he also said the new terms will not change how Meta crawls publishers in future, stressing that content owners can still manage how their content is accessed by using robots.txt. To Ars Technica, the spokesperson put it as: there is no change to how Meta uses web crawlers, and any suggestion to the contrary is false. Meta's developer pages still document how to block its crawlers with robots.txt.

The third point is what lawyers and Ars's side-by-side comparison found. Three legal experts reviewed the terms at Nieman Lab's request. Matt Topic, a media and intellectual property attorney at Loevy & Loevy, called it a remarkable provision to bury in terms of service, given that it would give Meta access to material that would likely cost many millions of dollars in a licensing market. James Grimmelmann, a professor of digital and information law at Cornell Law School, said that for news publishers this clearly represents permission for Meta to train AI on any content from their sites and possibly to deploy some of that content even in competition with the publishers themselves; asked about the alternatives, he said publishers either have to trust that Meta does not mean the full extent of what the terms say, or stop using business accounts. Benjamin Sobel, a law professor at the University of Wisconsin, likened it to tying in antitrust law: if you want to use the Meta products that require acceptance of the commercial terms, you have to also acquiesce to the scraping and crawling permissions exacted in those terms, comparing it to Google's policies that require consent to AI Overviews in order to appear in ordinary search. Ars also compared old and new text: the current commercial terms never mention test credentials, and that phrasing appears only in the Self-Service Ads Terms introduced this month, which let users grant Meta permission to scrape their content to generate ads; the new terms widen that from content connected to your ads to improving all of Meta's products.

The boundaries and open questions need equal clarity. First, the word AI does not appear in the clause; the training intent is inferred from a single use of train plus references to scrapers and crawlers. Second, there is no opt-out: the update reads that if you do not agree to the updated terms, please stop all access or use of the products for business or commercial purposes. Third, enforceability is unsettled: the experts said it is not clear whether a court would enforce the terms, but Topic's advice is that publishers should assume it would be enforceable. Fourth, precedence between robots.txt and the commercial terms is unresolved, and Meta did not answer whether test credentials would be used to bypass paywalls or other paid content; Grimmelmann said the wording technically allows Meta to sign up as a fake user to get past a publisher's registration wall. Fifth, the wider context is that fair use for AI training is still being litigated: publisher lawsuits against OpenAI are working through federal courts, including a May class action by five large book publishers alleging Meta pirated millions of books to train its Llama models. Sixth, the terms arrive alongside Meta One, a paid subscription that limits how many external links some accounts can post — Nieman Lab's summary being that Meta has begun charging users to post links to the same content it will be permitted to crawl for free.