Meta’s updated commercial terms, due to take effect October 30, authorize the company and its contractors to crawl websites associated with business accounts and use retrieved content to develop, improve, research and train its products. Publishers and legal specialists are questioning whether that language could be used to claim permission for AI training on linked news sites.
The terms cover domains, pages and URLs connected with a business’s products or services. They do not explicitly mention AI, but they refer to automated scrapers and to training. Businesses that reject the terms are directed to stop using Meta’s commercial services rather than offered a clause-specific opt-out.
Meta told Ars Technica that there is no change to how its web crawlers operate and called suggestions otherwise false. The company also said it has no plan to use the terms to ignore robots.txt instructions, and publishers can continue blocking its crawlers.
That leaves both intent and enforceability disputed. Legal experts cited in the report said a court may not accept such broad terms, particularly when the affected website content is separate from material posted to Meta’s platforms. News organizations are concerned because agreeing through a business account could complicate future licensing claims. The text creates a possible legal argument for Meta, not proof that it has begun new scraping or that the argument would prevail.