Meta has removed wording from its updated commercial terms that legal experts said could let it collect content from the websites of businesses that advertise or run pages on its platforms and use that material to train artificial intelligence models, TechCrunch reported Saturday. The terms bind every company with a Meta business account, so the episode is a reminder that the fine print businesses accept to buy ads can also govern what a platform may do with their own web content.
What Happened
Meta recently revised its Commercial Terms, the contract that applies when organizations use Facebook, Instagram and other Meta products for business purposes. According to TechCrunch, the new text said Meta could use any content from a business's domains, web pages or URLs to provide, develop, improve, research and train its products and services.
Legal experts interviewed by the journalism site Nieman Lab called it a "remarkable provision" and said it could allow Meta to scrape the sites of any publisher with a Meta business account and feed that material into AI training, TechCrunch reported. The Verge then reported that the language had been taken out.
Meta has disputed that the clause changed anything. A spokesperson said the updated terms did not reflect a change in how Meta crawls the web. In a later statement quoted by TechCrunch, Meta said, "the crawling terms we consolidated only applied to our ads and business products, and we've clarified that in the terms themselves," adding that the change "does not change how Meta uses web crawlers."
The English-language preview of the updated terms, which take effect Oct. 30, 2026, reflects that revision. Its "Licenses & Rights You Provide Meta" section still authorizes Meta to use automated software to access and crawl domains and web pages associated with a business's products or services, or reached through use of Meta's products. But the stated purposes are now to provide, develop, improve, research and maintain advertising and business products, and to maintain the integrity and security of Meta's products. The words "train" and "any content from" no longer appear in that paragraph. The current Commercial Terms have been in force since March 6, 2024.
TechCrunch did not say when the terms were first changed or when the language was removed.
Why It Matters
For boards and general counsel, the question is less what Meta intends than what its contract permits. The terms are accepted when a business sets up ad accounts or pages, yet their license language decides how the company's own site content may be used. A broad training grant would have given Meta rights that many publishers are trying to deny AI developers through crawler blocks, licensing deals and litigation.
The revised wording narrows the stated purposes, but it still grants Meta crawling access to business websites for advertising, business products and security work. Organizations that block AI crawlers on their sites may want to check whether that contract permission sits comfortably with their own policies.
Context
AI training on web content remains contested in court. Meta defeated a high-profile copyright suit over training models on books in June 2025, TechCrunch noted, while other cases against AI companies, including a publishers' suit against Meta reported in May, continue. Platform terms of service are an increasingly scrutinized channel through which content rights can pass to AI developers without a separate negotiation.
What to Watch
- Whether the final text that takes effect Oct. 30 matches the revised preview.
- Whether Meta publishes details of which crawlers access business sites and for what purposes.
- Whether publishers or regulators press other platforms to disclose similar crawling or training grants in their business terms.
