Meta has retreated from the terms change that alarmed publishers. The clause letting the company use content from domains associated with business accounts to train and improve its AI products no longer appears on its commercial terms page.
How it started: Nieman Lab reported the October 30 update covered websites tied to accounts using Meta Business Suite, the dashboard behind Instagram, Facebook, Messenger and WhatsApp channels. Legal scholars, including Cornell’s James Grimmelmann, warned it could be read as overriding a site’s robots.txt instructions and skipping a licensing payment.
Meta’s first answer was denial. Communications chief Andy Stone and spokesperson Daniel Roberts said nothing about web crawler behavior had changed. That held for several days as the story spread across outlets covering publisher economics.
The company then changed its position and its document. Roberts told The Verge the crawling language only ever applied to advertising and business products, that the text now says so plainly, and that robots.txt remains the mechanism for opting out. He added the update grants no new rights.
For publishers, the episode shows that public pressure can still move a platform’s contract language before it takes effect, at least when the clause is legible enough for lawyers to quote back.
