Meta Removes Term That Let It Scrape Website Data for AI Training After Backlash

Meta has removed a recently added rule from its commercial terms of use that suggested the company could scrape content from news and other websites using a Meta business account to train its artificial intelligence models. The deleted section stated that Meta could use any content from such web pages or URLs to develop, improve, research, train, and maintain Meta products and services. Legal experts, including those from Neiman Lab and Cornell law professor James Grimmelmann, pointed out that the provision would have allowed Meta to take free content from sites and bypass blocking instructions in a site's robots.txt file. TechCrunch also reported on the controversial update before the company reversed its decision. Following public criticism, a Meta spokesperson and Andy Stone, the company's vice president of communications, stated that the update did not signal any change in how Meta crawls the web. The company claimed the text only applied to ads and business products and that content owners can still use robots.txt to state their preferences.

Context

Major technology companies frequently update their terms of service to collect public data for artificial intelligence development, often leading to pushback from users and publishers over privacy and copyright concerns. Other firms, such as WeTransfer, have similarly rolled back policy changes in the past after facing public criticism over machine learning data usage.