Meta Drops AI Training Clause From Business Terms

On October 15, 2026, Nieman Lab revealed something buried in Meta's upcoming Commercial Terms: language that appeared to let Meta crawl any website "associated with" a business account and use what it found to train its AI products. Two days later, the most alarming word in that clause — "train" — had quietly disappeared. Here is what actually happened, what changed, and what still applies to your business from October 30.

What Meta's new terms originally said

Meta's updated Commercial Terms, scheduled to take effect on October 30, 2026, contained a new paragraph in Section 1. It authorized Meta "or any entity acting on our behalf" to use automated software — scrapers, bots, spiders, crawlers, and user agents — and even to create test credentials, in order to access, index, cache, analyze, collect, and crawl domains, pages, and URLs associated with a business's products or services, and to periodically pull updates from them.

The next sentence was the one that set off alarms. It said Meta could "use any content from [a business account's] domains, web pages, or URLs to provide, develop, improve, research, train, and maintain the integrity of" Meta products and services, as reported by TechCrunch on October 17.

Who would this affect? A very large group: anyone using Facebook, Instagram, or WhatsApp for commercial purposes — running ads, managing a Facebook Page, listing products, using Ads Manager or Meta Business Suite. The rewrite also made acceptance passive: continue using Meta's business products after October 30, and you agree. The only way to refuse was to stop all business use.

Why publishers and lawyers pushed back

Three legal experts who reviewed the terms for Nieman Lab said the wording could give Meta a contractual argument to train its AI systems on publishers' own websites — no license required. Matt Topic, a media and intellectual property attorney at Loevy & Loevy, said: "This is a remarkable provision to bury in terms of service, given that this would give Meta access to material that would likely cost many millions of dollars in a licensing market." He added that while it is unclear whether a court would enforce the clause, "publishers should assume that it would be enforceable."

James Grimmelmann, a professor of digital and information law at Cornell Law School, said: "For news publishers, this clearly represents permission to 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."

The News/Media Alliance, an industry group representing publishers, called the initial terms "outrageous," and Jason Kint, head of the Digital Content Next trade group, wrote that the terms reached "outside [Meta's] walled gardens."

Meta's first response came from spokesman Andy Stone, who said the clause was designed to "pull accurate information about products for sale on third-party websites when making on-platform ad and content recommendations, not about scraping news sites." Stone also said the update did not change how Meta uses web crawlers and that robots.txt preferences would still be honored — "This is not about scraping news sites."

What Meta changed — and when

Sometime over October 16–17, the word "train" vanished from the commercial terms page, and The Verge reported that the disputed provision had been removed. Meta issued a statement acknowledging the criticism: "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. It does not change how Meta uses web crawlers."

According to reporting that compared the two versions, the crawling permission now says Meta may pull content from linked domains "in order to provide, develop, improve, research, and maintain advertising and business products," plus "maintain the integrity and security of Meta Company Products." The effective date did not move: the page still reads October 30, 2026.

What did NOT change

This is the part many business owners will miss, so be precise:

• The terms still take effect October 30, 2026, and acceptance remains passive — continued business use counts as agreement.
• The broad crawling right stays. Meta can still use bots and crawlers on domains associated with your business, and can still create test credentials. Only the stated purpose narrowed.
• There is no opt-out short of stopping all business use of Meta's products.
• robots.txt is not mentioned in the contract. Meta says it honors robots.txt preferences, but how these terms interact with a site's own robots.txt has never been tested in court — an expert caution from the original reporting worth remembering.
• This is not a guarantee about AI training. The specific wording that read as an AI-training permission is gone, which weakens that legal reading, but Meta retains broad crawling rights and has separate AI training policies. The crawler reportedly involved, Meta-ExternalAgent, is a mixed-use agent rather than one dedicated to a single purpose.

What this means for your business: practical steps

1. Check your exposure. If you run ads, manage a Page, list products, or use Meta Business Suite, these terms are aimed at you.
2. Review your robots.txt. It remains the standard way to tell crawlers what they may and may not fetch. Make sure it reflects your intent for mixed-use crawlers like Meta-ExternalAgent.
3. Check your server logs. If you want to know whether Meta's crawlers visit your site, look for their user agents in your access logs — evidence beats assumptions.
4. Decide before October 30. After that date, continuing to use Meta's business tools counts as acceptance. There is no partial opt-out.
5. Watch for further edits. Meta revised this clause once under pressure; the terms page is the source of truth if wording changes again.

If you are a publisher worried about the bigger picture — AI systems using your content while search traffic shifts — you may also want to read our piece on how Google's AI Mode now gives publishers' links more visibility, and on Google's SynthID detector for spotting AI-generated images, since content provenance is becoming part of this fight. For a different Meta development this week, see Facebook's new Your Algorithm control for reshaping your Feed.

What remains uncertain

Honesty matters here: we do not know whether a court would read the revised terms as overriding a site's robots.txt — that question is legally untested. We also do not know how Meta's separate AI training policies apply to content its crawlers collect under these terms. And the underlying disputes continue: Meta defeated a high-profile copyright lawsuit last year, but other cases against AI companies, including Meta, are still ongoing.

Featured image: AI-generated illustration, not a real photograph.

Sources

• TechCrunch — Meta removes language that raised scraping and AI training concerns (opened and verified)
• StartupFortune — Meta quietly deletes AI training clause from its new business terms
• PPC Land — Meta's new terms let it crawl business websites to train AI from Oct 30