Meta's new business terms may still let it use news sites for AI

Meta has changed its new business terms to drop language about training AI on news sites. Legal experts say the edits may not be enough. The terms cover business accounts and take effect October 30th.

What changed

The first version said Meta "may use any content" from websites tied to a business account to "develop, research" and "train" its products. Experts told Nieman Lab this could let Meta argue its business terms override robots.txt. That is a file that tells scrapers, including search engines like Google, what they may crawl. After pushback from the publishing industry, Meta removed the word "train."

What experts still worry about

Matt Topic, a media and intellectual property attorney at Loevy & Loevy, said the new version still lets Meta use scraped content to develop any "business product." He noted the term does not seem to be defined, and he thinks "develop" is broad enough to include "train."

James Grimmelmann, a Cornell Law professor, said the change does nothing to limit the "associated with your products or services" language, which he said would let Meta scrape entire sites.

What Meta says

Meta spokesperson Daniel Roberts said content owners can use robots.txt to signal their preferences. He said the update "is not about scraping news sites" and does not give Meta new rights that supersede those preferences.

The material does not say how the terms will work in practice. Anyone with a website tied to a Meta business account should read the terms before October 30th and check their robots.txt settings.