Meta Drops the Word "Train" From Commercial Terms Taking Effect October 30: It Can Still Periodically Crawl Sites Linked From Business Pages, and a Law Professor Says "Develop" Is Broad Enough to "Drive a Truck Through"

What changed

Meta's new commercial terms of service for business users take effect on October 30. According to Ars Technica's October 20 report, Nieman Lab had earlier cited legal experts who pointed out that the crawling clause let Meta use "any content" from websites linked on business pages "to provide, develop, improve, research, train, and maintain the integrity" of any Meta product, service or feature. The concern was that Meta could make every organization with a business page, news organizations included, hand over AI training data by default and sidestep licensing fees.

After the criticism, Meta changed the wording. It removed "train" and "any content", and now lets Meta "periodically pull updates" from business websites to "provide, develop, improve, research, and maintain advertising and business products", as well as to maintain the integrity and security of its products. A Meta spokesperson told Ars that the consolidated crawling terms only apply to ads and business products, that they don't change how Meta uses web crawlers, that content owners can still use robots.txt to signal their preferences, and that "this is not about scraping news sites." Meta also confirmed that it didn't notify users of the change.

Why experts still aren't satisfied

Several legal scholars who spoke to Nieman Lab weren't persuaded. Media and intellectual property attorney Matt Topic noted that the revised terms still let Meta use scraped content to develop any "business product", a term that isn't defined, while Meta is plainly developing AI business products: "develop" is broad enough to include "train." Cornell Law School professor James Grimmelmann said the added language on permitted uses was an improvement, but "develop… advertising and business products" is "still broad enough to drive a truck through." University of Wisconsin law professor Benjamin Sobel suggested the terms explicitly commit to honoring robots.txt. Compliance with robots.txt is voluntary, so leaving it out of the terms leaves room to ignore it.

What it means for content owners

The episode shows that whether the word "train" appears in the terms has become a focal point between platforms and content owners, but what actually decides how data can be used is broad language like "develop" and "business products", and whether robots.txt is binding. For media and businesses with Meta business pages, the main way to limit crawling is still robots.txt and server-side blocking of Meta's crawlers on your own site, while watching whether the terms change again.