Illustration of a web crawler scanning a news website homepage

Meta Drops AI-Training Language From New Business Terms After Publisher Backlash

ByChris Wade

Key takeaway: Meta has removed controversial wording from its updated Commercial Terms that legal experts warned could let the company scrape news publishers’ websites to train AI models. The clause takes effect October 30, 2026, and will now state that crawling applies only to ads and business products. Meta insists its actual crawling practices are unchanged.

What Meta’s new terms originally said

Meta recently updated its Commercial Terms of use — the contract for businesses using Facebook, Instagram, and WhatsApp commercially — with a clause declaring that the company can “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, according to TechCrunch. Because Meta’s own definition of its products includes its AI services, the word “train” read, to critics, as a license to train AI models on scraped website content.

The updated terms take effect on October 30, 2026, and per reporting on the terms text, acceptance is passive: continuing to use Meta’s commercial products counts as agreement, with no partial opt-out — businesses that disagree may “stop all access or use” of Meta’s products for commercial purposes.

Why publishers and legal experts pushed back

The wording was first surfaced by Nieman Lab on October 15. Legal experts interviewed there called the clause a “remarkable provision” that could give Meta permission to scrape the website of any publisher with a Meta business account and use that content to train AI models, as reported by TechCrunch.

The concern was practical: most major news publishers keep Meta business accounts to run ads, manage Facebook Pages, and promote stories. Under the broad reading of the clause, that ordinary business presence could place a publisher’s own website within Meta’s crawling reach — without any licensing negotiation. The dispute lands amid a wider fight over AI training data: Meta defeated a high-profile copyright-infringement lawsuit last year, but other cases against AI companies, including Meta, continue.

What Meta changed — and what stayed the same

After several days of criticism, Meta edited the terms, as first reported by The Verge and covered by TechCrunch on October 17. In a statement, the company said:

“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.”

The revised terms now specify that automated website crawling is intended to “develop, improve, research and maintain” advertising and business products — dropping the broad “train” language, according to Anadolu Agency.

What did not change: the crawling authorization itself. Meta can still use scrapers, bots, and crawlers to collect content from websites tied to businesses using its platforms. Spokesperson Daniel Roberts said the changes do not alter Meta’s existing crawling practices and that website owners can continue using robots.txt to signal their access preferences.

What this means for publishers and businesses

If you run a business account on Meta’s platforms, three things are worth checking before October 30:

  • Review the website linked to your Meta business account. The crawling terms apply to domains and URLs associated with your products or services.
  • Check your robots.txt. Meta says it continues to honor robots.txt signals — make sure yours says what you intend about Meta’s crawlers.
  • Weigh the trade-off. There is still no way to keep your Meta business presence and remove your site from the crawling terms; the only opt-out is to stop using Meta’s commercial products.

Analysis: a real edit, not a full reversal

Analysis — this section is USABite’s interpretation, not a reported fact. Meta’s edit narrowed the legal hook, but it did not add a guarantee. The company removed the single word — “train” — that made the clause read as an AI-training license over other people’s websites. But the crawling authorization, the October 30 date, and the passive-acceptance structure are all still in place.

The speed of the walkback is telling: Nieman Lab surfaced the language on October 15, and Meta had edited it out within about two days. That suggests the company saw genuine legal and reputational risk in leaving the wording as-is — which, in turn, suggests publishers should not treat the revised terms as a clean bill of health.

FAQ

Did Meta change its web crawling practices?

No, according to Meta. The company says the revised Commercial Terms clarify that its crawling provisions apply to advertising and business products, and that the changes “do not change how Meta uses web crawlers.” Spokesperson Daniel Roberts added that website owners can continue to use robots.txt to indicate their access preferences.

When do Meta’s updated Commercial Terms take effect?

October 30, 2026. The revised terms say that automated website crawling is intended to develop, improve, research, and maintain advertising and business products. Businesses that continue using Meta’s commercial products after that date accept the terms.

Could Meta use publisher websites to train AI models?

The original wording worried publishers and legal experts because it allowed Meta to “train” its products using content from any business account’s websites, and Meta’s products include its AI services. Meta has now removed that language and says the crawling terms were only ever meant for ads and business products — but independent lawsuits over AI training data continue.

Can I opt out of Meta’s website crawling?

There is no partial opt-out. Per the terms, businesses that do not agree can stop using Meta’s products for business or commercial purposes. You can still use robots.txt to signal crawling preferences, and Meta says it continues to honor those signals.

Sources


Similar Posts