Meta has pushed back the start date of its new business terms by three weeks, from October 30 to November 20. The change covers the Commercial Terms and the Self-Serve Ad Terms — the contract every business accepts by running a Facebook page or buying ads through Ads Manager. The reason: publishers and advertisers complained to antitrust regulators in Brussels and London.
The dispute is about a clause that lets Meta use automated software — scrapers, bots, spiders, "user agents" — to access domains "associated with your products or services", and to index and cache what it finds in order to "provide, develop, improve, research and maintain advertising and business products". Publishers read that as a right to use their websites and archives for AI. Meta rejects the reading.
The wording has already been tightened once. A standalone sentence permitting Meta to use such content to "train" its products disappeared from the published text; the crawling authorisation now ends with a stated purpose. The News/Media Alliance says that changes little, because "advertising and business products" is undefined and the clause still covers any URL associated with a publisher's products — which, for a subscription business, can mean the whole site. The permission to create test credentials and get past registration walls also survives. And although a Meta spokesperson told Reuters that robots.txt preferences would still be honoured, that commitment is not in the contract text.
Just as notable is Section 5, "AI Features". It states that AI features "may make mistakes" and that the business is "solely responsible for and assume[s] any and all risks associated with such AI features and any resulting Content". That covers not only tools a business chooses to use but automatically applied changes to ads — sitelinks, translations, generated images. Section 6a still requires businesses to indemnify Meta against claims "of any kind"; Section 9g still lets Meta "modify, limit, suspend or discontinue any commercial products, services and features at any time without prior notice or liability".
How consent works matters more than the date. Under Section 7, continuing to use Meta products for business purposes after notice means accepting the terms. There is no partial consent: a publisher cannot accept for its Facebook page and refuse for its website. It can leave Meta, or it can be bound. The delay moves the moment of deemed acceptance; it does not change the mechanism.
Nor is the delay fully visible yet. On the evening after the announcement, Meta's UK preview page still displayed October 30 as the effective date. The European Commission welcomed the postponement and used the occasion for a separate reminder: it has sent requests for information to companies that have not yet published summaries of the data used to train their AI models, an obligation under the EU AI Act. The European Publishers Council wants a suspension rather than a pause. The News/Media Alliance had given Meta until October 26 to respond.
Three weeks is not much time for any of those processes, which is why publishers are treating November 20 as a deadline rather than a resolution. Every brand with a page, a catalogue and an ad budget is on the same clock.
