Meta drops terms that critics said let it train AI on publisher sites

Meta has removed a line from its commercial terms that critics read as permission to scrape websites and use the content to train its AI. The Verge reported the change on October 17, 2026, and TechCrunch followed the same day.

The removed text said Meta could use any content from the domains, web pages or URLs tied to a business account to build, improve, research and train Meta products. Nieman Lab and legal experts, including Cornell law professor James Grimmelmann, warned it could let Meta train on publishers' sites for free and even override the blocking rules a site sets in its robots.txt file.

Meta had already denied that the terms changed how its web crawlers work. After the backlash, it took the language out entirely. In a statement to The Verge, a spokesperson said the consolidated crawling terms only covered Meta's ads and business products, that content owners can still use robots.txt to signal their wishes, and that the update gives Meta no new rights over those preferences.

Why it matters

Anyone running a site with a Meta business account should know that robots.txt is still the main tool for controlling AI crawlers, and that platform terms can quietly change.

Source: The Verge