Meta has updated its commercial terms, removing language that previously allowed the company to scrape external websites for AI training purposes. The revision follows significant criticism from publishers and legal experts who argued the original wording could bypass paywalls and undermine existing licensing agreements.
The original terms, which were set to take effect on October 30, 2026, authorized the use of automated tools to collect data from URLs linked to business accounts. Following backlash, Meta clarified its policy around October 16–17, 2026, restricting scraping permissions strictly to advertising and business products.
The company also confirmed it will respect standard web protocols, including robots.txt. This change addresses concerns from industry groups like the News/Media Alliance, which had labeled the initial terms as a potential abuse of market power.
Meta currently maintains paid AI licensing deals with media outlets such as CNN and Fox News. Legal observers noted that the initial terms could have allowed the company to scrape content that it otherwise pays millions of dollars to license, creating a conflict with its own business practices.
