AitopkitMeta has deleted a controversial clause from its commercial terms that legal experts said could have given the company permission to scrape publishers' websites and use their content to train AI models. The deletion came on October 17, after days of escalating criticism from news publishers, trade groups and copyright lawyers.
The original language — flagged first by Nieman Lab on October 15 — said Meta could use "any content from such domains, web pages, or URLs to provide, develop, improve, research, train, and maintain the integrity of" its products. The terms covered any business promoting content across Facebook, Instagram, WhatsApp or Messenger, and were set to take effect October 30 with no opt-out other than stopping business use of Meta's products entirely. Cornell digital-law professor James Grimmelmann called the language "absurdly broad," and publishers warned it could override robots.txt — the standard signal sites use to tell crawlers what they may copy.
Meta insists nothing of substance changed. Meta communications representatives said the crawling terms only ever applied to its ads and business products, that the update gave Meta no new rights over robots.txt signals, and that it "does not change how Meta uses web crawlers." The word "train" quietly disappeared from a preview revision late Friday, October 16, before the whole clause was removed on Saturday.
The episode lands in an already tense week for Meta's training-data practices — and alongside a broader pushback against AI companies' scraping: Yelp, Delta and major banks are moving to block AI agents from OpenAI, Meta and xAI, while Anthropic's agents were found exploiting live websites. For publishers, the message is that reading terms of service carefully now works — and Meta has learned that a single word can start a fight.