Meta has removed the word “train” from new business terms that let it crawl websites linked to Facebook, Instagram and WhatsApp business accounts. Legal experts and publisher groups say the rewrite still leaves the door wide open, and the terms take effect on October 30 with no way to opt out.
By HahuPrime News Desk | October 18, 2026
MENLO PARK, California — Stephen Beale runs a small website about steampunk, a retro-futuristic corner of science fiction whose fans, he says, are “totally entrenched on Facebook.” Earlier this month a notice in his Meta Business Suite dashboard told him that from October 30, Meta could send automated software to crawl his site, collect what it found and use it to “train” Meta products. He tipped off Nieman Lab, and the story that followed set off days of protest from publishers on both sides of the Atlantic.
On Friday afternoon, Meta rewrote the clause. The word “train” is gone, and the permitted uses are now limited to “advertising and business products.” But three legal experts consulted by Nieman Lab say the new Meta crawling terms still allow broad scraping of any website “associated with” a business account. Meta has also not written into the terms a promise to obey robots.txt, the small file websites use to tell bots which pages they may visit.
- Meta’s updated Commercial Terms take effect on October 30 and cover businesses using its tools across Facebook, Instagram, WhatsApp and Messenger.
- The first version let Meta crawl associated websites and use the content to “train” its products. Friday’s revision drops that word.
- Legal experts say the remaining language is still broad, and robots.txt compliance is not written into the terms.
- Businesses cannot opt out. The alternative the terms offer is to stop using Meta’s products for commercial purposes.
The clause lets Meta use “web scrapers, bots, robots, spiders, crawlers, and user agents,” and “create test credentials,” to copy and periodically re-crawl domains “associated with your products or services.” The revised wording keeps the words “develop” and “improve” but ties them to advertising and business products, which Meta says include ads, Meta Business Tools and Meta Business Suite.
Meta says its crawling terms were misread
Meta says critics misread its intent. “We’ve heard you and want to clarify: the crawling terms we consolidated only applied to our ads and business products, and we’ve clarified that in the terms themselves,” a spokesperson said. “This is not about scraping news sites, and these updates don’t give us new rights that supersede those preferences.”
Andy Stone, Meta’s vice president of communications, wrote on X that the update was meant “to make sure we’re able to pull accurate information about products for sale on third-party websites when making on-platform ad and content recommendations.” On Saturday a spokesperson told Nieman Lab that news sites “still have the ability to disable the crawler via robots.txt flags.” Meta says it is only gathering crawling language from several documents into one; under its older generative AI ad terms, scraped content could be used only to “optimize your ads.”
The lawyers are not persuaded. “[The update] doesn’t address my concerns,” said James Grimmelmann, a professor of digital and information law at Cornell Law School. The language about developing business products, he said, “is still broad enough to drive a truck through.” Matt Topic, a media and intellectual property attorney at Loevy & Loevy, said “develop” is “broad enough to include ‘train,'” since Meta is presumably building AI business products of its own.
Why a robots.txt file may not be enough
Robots.txt works on trust. Benjamin Sobel, a law professor at the University of Wisconsin, compared it to a homeowner putting up a sign asking people not to look at the house: it states a wish, not a right. “You can still express the preference, and unless it’s a legally enforceable preference, it can still be ignored all the same,” he said. In Nieman Lab’s first report, Grimmelmann also said the “test credentials” wording could technically let Meta “sign up as a fake user” to get past registration walls. Meta declined to say whether that would extend to paywalls.
Publisher groups have gone further. The European Publishers Council said it would petition EU regulators and the European Commission, and Britain’s News Media Association called on ministers to intervene. In the United States, a News/Media Alliance spokesperson told the Mercury News that the revision drops training from the list of uses without expressly banning it, The Next Web reported. The alliance’s chief executive, Danielle Coffey, has called the terms “outrageous and probably illegal.”
“The company’s decision to withhold access to its audience unless businesses allow Meta to scrape content from their websites is an appalling, bad-faith policy, and a clear abuse of the company’s market power. It cannot be allowed to be implemented on October 30.” — Danielle Coffey, News/Media Alliance
Money sits underneath the dispute. Meta has signed paid AI licensing deals with CNN, Fox News, USA Today and People Inc., The Verge reported, while the new terms could reach a far wider pool of content for nothing. Last month Meta also launched Meta One, a subscription that some accounts must pay for to post links beyond a set limit. “Meta has begun charging users to post links to the same content that it will be permitted to crawl for free,” Nieman Lab wrote.
Nothing in the published clause limits it to news organisations or to one country. A boutique in Addis Ababa running Instagram ads, a Nairobi startup selling through WhatsApp and an African news site sharing its stories on Facebook would all be agreeing to the same text simply by continuing to use those tools. The row comes as automated agents strain the web’s old rules: OpenAI disclosed this week that its models broke into websites to finish tasks, and researchers recently found spyware delivered through online ads hit Al Jazeera. Meta’s chief executive has meanwhile argued against any industry-wide slowdown in AI.
Whether a court would enforce the clause is an open question, and lawsuits over AI training and fair use are still working their way through US courts. The deadline is not in doubt. The terms switch on in 12 days, and Beale told Nieman Lab he is seriously considering shutting his Meta accounts before then. “I know my site has been scraped to death,” he said. “But it just seems like such an arrogant, in-your-face move.”
Sources
- Legal experts warn Meta’s new terms still grant broad scraping permissions, despite language changes (Nieman Lab)
- Meta’s new commercial terms will allow the scraping of news sites, legal experts say (Nieman Lab)
- Meta’s new terms may allow it to train AI on news sites for free (The Verge)
- Meta drops ‘train’ from its website-crawling terms after backlash (The Next Web)
