Meta walks back commercial-terms wording after publisher scraping backlash

What changed in Meta's commercial terms

Meta updated the commercial terms that govern how its automated systems interact with websites connected to a Meta business account. An initial draft of the terms, due to take effect October 30, contained language that legal experts and news publishers warned could let Meta use web crawlers to collect content from "virtually all news" publishers and others with a business presence on the platform, then train its AI models on that material. Following the backlash, Meta removed the contested wording. The revised terms now limit the stated purpose of crawled content to "advertising and business products" rather than all Meta products.

Why publishers flagged the original clause

Industry coverage described the original draft as effectively giving Meta a permission slip to pull content from any site tied to a Facebook or Instagram business page. Nieman Lab's reporting, cited across the coverage, noted that virtually all news publishers operate such pages, which would have exposed their archives to scraping by default. Trade outlets framed the move as an escalation of a "pillaging" pattern in commercial-to-AI data harvesting, with one outlet warning the language "could allow website scraping and the use of publisher content" if left in place.

What the revision actually does, and what it does not

The clause that drew the most criticism — explicit permission to train AI on scraped publisher material — no longer appears in the revised terms. Reporting on the Friday publication describes the change as Meta dropping "train" from the scraping clause while keeping broader crawling rights intact. In other words, Meta retains the technical ability to crawl business-linked sites, but its commercial terms no longer cite AI model training as a permitted use, and the stated scope is now narrower.

Publishers, agencies and site owners preparing for the October 30 effective date are being advised to read the updated language carefully and to confirm what their existing Meta business settings allow.

Reading the new language before October 30

Because Meta's revised terms take effect October 30, the practical workload falls on writers, editors and publishers who run business accounts. Anyone responsible for the text a company publishes online should compare the current commercial-terms document against earlier drafts, mark any clauses that still permit crawling or reuse, and decide whether their bot-control files, HTML Cleaner pipelines and publishing workflows need adjustment before the effective date. Doing that review before the deadline keeps content, metadata and crawl permissions aligned with the narrower scope Meta has now stated.

What practitioners can do next

A reading list for teams that publish through Meta surfaces naturally from this story. Site owners running on Meta business accounts can audit which URLs are linked to those accounts, verify that any opt-out mechanisms — including robots rules — are still respected, and refresh their content-ingestion and cleanup tooling, including steps such as running archived pages through an HTML Escape Large Text review or stripping stray glyphs via a Special Characters Remover.

Editorial teams should also confirm that their AI-text detection and watermarking checks, described in related coverage of model-output markers, still function when synthetic content is mixed with material the team has rewritten rather than generated. These are tasks a practitioner can complete before October 30 to ensure the new commercial terms do not catch their workflow off guard.