Meta Could Still Scrape Sites for AI Training, Legal Experts Warn

Terms Changed After Backlash

Meta’s revised business terms no longer say that the company will train its products using scraped content. The changes followed strong opposition from the publishing industry. The terms take effect October 30th. An earlier version said Meta could use content from websites associated with a business account to develop, research and train its products.

Experts See Broad Language

James Grimmelmann, a Cornell Law School professor, told Nieman Lab that the revision leaves intact wording that could allow Meta to scrape entire sites. He said the terms still refer to content “associated with your products or services.” Matt Topic, a media and intellectual property attorney at Loevy & Loevy, said the revised terms permit use of scraped content to develop any “business product.” He argued that “develop” could include training, and noted that the term business product is not defined.

Meta Defends Its Position

Meta has said the terms are not intended to let it scrape news content for free. Spokesperson Daniel Roberts told The Verge that site owners can use robots.txt to communicate their preferences.

Roberts said the update was not about scraping news sites and did not grant Meta new rights that override those preferences. Earlier, legal experts had warned that Meta might argue its commercial terms superseded blocking instructions such as robots.txt.