Meta has removed the word ‘train’ from commercial terms due to take effect on 30 October, but lawyers say the remaining permission to develop products could still cover AI training, The Verge reported on 22 October. The change follows criticism of terms governing business accounts that appeared to permit Meta to use content from websites associated with those accounts, including news sites.
Key points
- Meta removed ‘train’ from the terms after criticism from publishers, The Verge reported.
- Two lawyers said the wording could still permit Meta to scrape sites or use their content to develop AI products.
- A published preview checked on 18 October still contained the sentence permitting training, PPC Land reported.
- Meta says the crawling provisions apply to ads and business products and that site owners can signal preferences through robots.txt.
Meta’s permission to develop products
The earlier version of the terms said Meta could use content from sites associated with a business account to develop, research and train its products. Publishers objected to the potential use of news content for AI training. AI Affairs previously reported on the permissions in Meta’s 30 October terms for business-linked websites.
James Grimmelmann, a professor of digital and information law at Cornell Law School, told Nieman Lab that Meta’s revision left in place wording about sites “associated with your products or services”. He said that wording could let Meta scrape entire sites. Legal experts had also told Nieman Lab that Meta might argue the commercial terms took precedence over robots.txt instructions, which website owners use to direct crawlers.
Matt Topic, a media and intellectual property attorney at Loevy & Loevy, told Nieman Lab that the revised terms still let Meta use scraped material to develop a business product. “I think ‘develop’ is broad enough to include ‘train,’” he said. His reading concerns what the contract could permit, rather than a change to Meta’s crawler activity.
The 18 October Meta terms preview
Meta had removed the disputed training sentence from its commercial terms page, The Verge reported. A different version was visible in the published preview of the updated terms on 18 October: the sentence allowing Meta to use content to “train” its products remained there, PPC Land reported after checking the UK English page.
The preview also authorised Meta or an entity acting on its behalf to use automated tools to access, collect and crawl domains, pages or URLs associated with a business’s products or services. It permitted the creation of test credentials and periodic collection of updates from those sites. PPC Land found that the relevant provisions matched the version it had used for an earlier report on 16 October.
The preview’s crawling paragraph contained no reference to ads or business products, PPC Land reported. It also made no mention of robots.txt. Those findings concern the preview PPC Land checked on 18 October, while The Verge’s account of Meta’s removal was published on 22 October.
Roberts describes Meta’s crawler policy
Meta spokesperson Daniel Roberts said the consolidated crawling terms applied only to the company’s ads and business products. Content owners could use robots.txt to signal their preferences, he said, and the update did not give Meta rights that overrode them. Meta vice president of communications Andy Stone said the update did not change how the company used web crawlers.
The preview checked by PPC Land said continued commercial use after notice of an update constituted agreement to the new terms. Its Section 7 said businesses that did not agree could stop using Meta’s products for commercial purposes. The preview gave 30 October 2026 as the date the updated terms would take effect.
