Meta Delays New Business Terms Amid Publisher Pushback Over AI Training

Meta has said it may delay its new business terms past November 20, after publishers objected that the terms could be used to train AI systems on content from Facebook and Instagram business pages. The Verge reported that the company may now delay the rollout indefinitely.

The updated terms were originally due to take effect on October 30. Meta first moved implementation to November 20 and is now weighing a further, open-ended delay. Until the updated terms take effect, Meta said its existing terms remain in place.

The delay covers Meta's updated Commercial Terms and Self-Serve Ad Terms, the agreements that govern business activity and paid advertising on the platform. Reuters reported that publishers and advertisers criticized the changes, and Meta said it is delaying implementation to address those concerns.

Meta has also revised the business terms to remove language suggesting it could train its products on content scraped, meaning automatically copied, from Facebook and Instagram business sites. The revision narrows what the earlier wording appeared to permit, but on its own it does not settle the question critics have raised.

Andy Stone, Meta's head of communications, said the company does not plan to make policy changes without full consultation with industry groups and leaders. If that consultation extends beyond November 20, he said, Meta would delay implementation of the updated terms. The rollout date is therefore tied to the outcome of those discussions rather than to a fixed calendar date.

The political response has been direct. Sen. Marsha Blackburn said Meta's revised terms indicate the company is attempting to exploit American creators and businesses. Sen. Adam Schiff urged Meta to reverse its business-term revisions.

Legal experts said the new terms could be used to freely scrape news sites. Nieman Lab covered that assessment in its reporting on the delay.

The broader context here is that the dispute is being fought in contract language. Terms of service and commercial terms define what a platform may do with the material its users and advertisers supply. That permission layer is separate from the technical controls a site uses to restrict crawling, the automated programs that copy pages. For engineers and data teams, the operative question is which text governs in the interim, and whether the eventual wording leaves room for the broad collection the legal experts describe.

In my view, the sequence so far, with a revision, a delay, a consultation and a possible open-ended postponement, suggests the final text will be shaped by the discussions Stone described rather than by the original draft. Publishers and advertisers have already signaled that the current language is unacceptable to them, and political attention adds pressure from a second direction. Whether the eventual terms draw clear lines on AI training and scraping will matter to anyone who builds on Meta's ad and business tools.

The case for a slower, more consultative process rests on the possibility of clarity. Terms that publishers, advertisers and developers can read without ambiguity give businesses something to plan around, and that would make the platform easier to build on.