Meta may use new commercial TOS to avoid millions in AI licensing fees

Meta has updated its commercial terms to require business accounts to authorize automated access to any domains, pages, or URLs they link to from Meta products, allowing Meta (or agents) to crawl, index, cache, analyze, and periodically pull updates from those sites and to use that content to develop, improve, research, and maintain its products and services. The new language explicitly permits creation of “test credentials” to retrieve content associated with a business’s products or services, and the terms are presented as non‑negotiable for businesses that want to remain on Meta’s platforms. Meta says the change consolidates crawling-related language from other policies and that content owners can still use robots.txt to control access.

Legal experts and publishers warn the update could be a major AI licensing power grab: by contractually securing broad scraping rights from businesses, Meta might avoid paying licensing fees for material used to train models or otherwise exploit publisher content, especially if courts later deem such training fair use. Critics say “test credentials” could enable bypassing paywalls and that bundling broad scraping consent into commercial agreements resembles tying, raising antitrust and competitive concerns. Meta disputes plans to ignore robots.txt, but publishers and lawyers worry the terms materially expand permissions beyond past, narrower ad- and seller-specific scraping provisions, prompting some publishers to consider leaving the platforms.