Meta’s revised commercial terms of use, set to take effect on October 30, authorize the company to employ automated software such as web scrapers to access, index, and collect content from domains “associated with” business accounts on its platforms, according to the source text. The update applies to any entity that manages a Facebook page, Messenger account, Instagram profile, or WhatsApp channel through a Meta business account, a category that, as Nieman Lab notes, virtually every news organization possesses—including Nieman Lab itself.
The language explicitly permits Meta to use the harvested content “to provide, develop, improve, research, train, and maintain the integrity of” its products and services. Elsewhere in its terms, Meta defines “AI at Meta” as one of those products, leading three legal experts consulted by Nieman Lab to conclude that the provision could be interpreted as granting Meta permission to scrape news sites for AI training. Matt Topic, a media and intellectual property attorney at Loevy & Loevy, described the clause as “a remarkable provision to bury in [the] terms of service” and warned that, while enforceability is uncertain, publishers should assume it would hold up in court. James Grimmelmann, a professor of digital and information law at Cornell Law School, added that the terms effectively allow Meta to train AI on any content from publishers’ sites and potentially deploy that content in competition with the original outlets, possibly building a comprehensive archive for Meta’s own use.
Benjamin Sobel, a law professor at the University of Wisconsin, characterized the arrangement as a form of “tying”: to continue using Meta’s business products, users must accept the accompanying scraping permissions. He likened the move to Google’s recent requirement that publishers consent to AI Overviews if they wish to appear in traditional search results.
The updated terms provide no opt‑out mechanism for business accounts before the October 30 deadline; Meta’s notice states that disagreement requires ceasing all business or commercial use of its products. A Meta spokesperson told Nieman Lab that the scraping language was previously scattered across various terms and has now been consolidated, and that publishers can still manage crawler access via robots.txt. However, the experts warned that the new commercial terms could be argued to supersede robots.txt directives, and that the ability to create “test credentials” might enable Meta to sign up as a fake user to bypass registration walls—though Meta declined to clarify whether such credentials could circumvent paywalls.
Legal scholars emphasized that any AI training using scraped content remains contingent on unresolved fair‑use questions, noting ongoing lawsuits such as the May class‑action filed by five large book publishers alleging Meta pirated millions of books to train its Llama language models.
The policy arrives shortly after Meta launched its paid Meta One subscription, which limits free external link posting on Facebook and Instagram unless users subscribe. Press Gazette reported that some news publishers are currently exempt from that link limit, though Meta did not comment on the exemption. Nieman Lab learned of the term change through a reader tip; Stephen Beale, founder of The Steampunk Explorer, said he is considering shutting down his Meta business accounts before the effective date, describing the move as “arrogant” and “in‑your‑face.”