Meta Removes Controversial Terms That Suggested Web Scraping for AI Training

Meta has revised its commercial terms to eliminate controversial phrasing that suggested the tech giant could scrape content from third-party websites, including news publishers, to train its artificial intelligence systems.

The move follows pushback from legal experts and digital publishing observers who raised alarms over recently consolidated commercial terms.

The Disputed Terms

The issue centered on specific phrasing previously published on Meta's commercial terms page:

"... Meta may use any content from such domains, web pages, or URLs to provide, develop, improve, research, train, and maintain the integrity of, the Meta Company Products, services, and features provided under these Commercial Terms and any other applicable supplemental terms."

Industry observers, including Nieman Lab and Cornell digital and information law professor James Grimmelmann, interpreted the clause as an effort by Meta to grant itself broad rights to harvest online content for AI training at no cost. Critics also argued that the wording could potentially override instructions set in a website's standard robots.txt file, which site owners use to signal whether automated web crawlers are allowed to access and index their pages.

Meta Responds and Clarifies

Meta leadership denied that the update altered its actual scraping practices. Andy Stone, Meta's vice president of communications, and spokesperson Daniel Roberts rejected claims of policy shifts, asserting that any suggestion that Meta had altered how it uses web crawlers was false.

Following the backlash, Meta opted to scrub the controversial language completely from its terms page. In an updated statement, Roberts clarified the company's position and the scope of the document:

"We’ve heard you and want to clarify: the crawling terms we consolidated only applied to our ads and business products, and we’ve clarified that in the terms themselves. It does not change how Meta uses web crawlers. As always, content owners can use robots.txt to signal their preferences. This is not about scraping news sites, and these updates don’t give us new rights that supersede those preferences."

The revision highlights ongoing tensions between major artificial intelligence developers and web publishers over copyright, data usage rights, and the extent to which existing online content can be leveraged to build and refine generative AI models.