Meta quietly pulled language from its business terms that suggested it could scrape publisher and news sites for AI training without permission.
After reports from Nieman Lab and others flagged that Meta's updated business terms could let it crawl and train on content from sites 'associated' with businesses using its platforms, including news publishers, Meta removed the contested language. A company spokesperson maintains the change doesn't reflect any actual shift in how Meta crawls the web, framing it as a clarification rather than a retreat.
The episode follows a familiar pattern: broad terms-of-service language quietly expands data rights, publishers and watchdogs catch it, and the company walks it back publicly while insisting practice hasn't changed. It's the second high-profile scraping-terms reversal from a major platform this year.
Publishers and any business with a web presence need to actively monitor platform terms changes rather than trust default policies — scraping rights are being tested incrementally through fine print, not splashy announcements. Expect continued friction between AI labs needing training data and content owners trying to gate it.
The daily signal, curated. Get it in your inbox.
Subscribe on LinkedIn →