Meta removes language that raised scraping and AI training concerns

Meta has taken out a provision from its recently updated commercial terms of use after the language drew scrutiny over web scraping and AI training. The original wording said the company 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. The Verge reported the clause has since been removed.
Legal experts quoted by Nieman Lab described the language as a remarkable provision, raising the possibility that it could authorize Meta to scrape the website of any publisher holding a Meta business account and use that material to train its AI models. The change arrived at a sensitive moment for the company: while Meta won a high-profile copyright infringement lawsuit last year, other cases against AI companies, including Meta, are still ongoing.
A Meta spokesperson maintained throughout that the updated terms did not represent a change in how the company crawls the web. In a statement about the removal, the company said it had heard the concerns and wanted to clarify that the crawling terms it consolidated applied only to its ads and business products, and that this has now been made explicit in the terms themselves. Meta added that the revision does not alter how it uses web crawlers.
For developers and product teams, the episode is a reminder that broad grants buried in commercial terms can travel well beyond their original context. The disputed clause sat in a document aimed at businesses running ads and other commercial products, yet its wording covered any content on a business's domains and URLs, which is exactly the kind of phrasing that scraping and training questions turn on. When terms describe both operational uses and model training in one list, readers have to assume the widest reading until the company narrows it.
The underlying tension is unlikely to disappear with this edit. Meta's crawler behavior is governed by how its systems treat robots.txt directives and other signals, and the company's position is that nothing about that behavior changed. That leaves open, as a matter of public record, how training data is sourced and what recourse publishers have when they object. The removal addresses the contract language, not the practice.
Why it matters: Anyone running a site tied to a Meta business account should treat platform terms as a live input to their data strategy, since wording can shift and be retracted quickly. The practical takeaway for product and legal teams is to track the clauses that mention training, not just usage, and to keep crawl directives current. Whether the retraction signals a broader tightening of how Meta handles training data is an open question, and my own inference is that scrutiny of such clauses will continue as litigation proceeds.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
A Meta spokesperson insisted that the updated terms did not reflect a change in the way Meta crawls the web.