Meta business terms and AI scraping
Where it stands: Meta's revised business terms drew concern because they appeared to permit scraping websites associated with its business users for AI development, with an October 30 effective date. Meta removed the scraping language in two passes and then delayed enforcement to November 20. An expert said the change did not restrict language that could let Meta scrape entire sites. A later clause would require business customers to indemnify Meta for claims tied to its AI features, including attorneys' fees; watch whether that indemnity survives to November 20.
Latest
- 2026-10-28 (~12:15 PM ET, The Verge — theverge.com/tech): second slip signal — Meta communications head Andy Stone says the company does "not plan to make policy changes without a full consultation with industry groups and leaders," signaling implementation may again push past the November 20 enforcement date amid continued publisher backlash over AI-training concerns. Watch: whether November 20 holds or slips again, the final operative terms text, a named publisher test case, regulator/AG attention.
- 2026-10-24 (~1:00 AM ET, StartupFortune — startupfortune.com): a liability clause layer — Meta's new business terms reportedly require customers to indemnify Meta for claims tied to its AI features, including attorneys' fees, shifting exposure for the AI's mistakes onto customers. Watch: whether the indemnity survives to the displaced November 20 enforcement date, business-customer reaction, and expert evaluation of the liability allocation.
- 2026-10-23 (~10:47 AM ET, The Verge — theverge.com/tech): development — Meta is delaying enforcement of its new business terms after publishers raised concerns the terms could allow scraping of news sites for AI training (earlier reported by Fast Company); spokesperson Daniel Roberts says they take effect November 20 — "to allow more time for us to explain to our partners and stakeholders what's changing and what's not" — displacing the October 30 effective date the thread carried, and following the October 22 language removal within a day. Watch: whether November 20 holds, the final operative terms text, publisher reaction and a named test case, regulator/AG attention.
Claims to watch · 1 (1 pending, 0 slipped, 0 hit, 0 missed)
- Meta, Oct 16, 2026: business terms with scraping permission effective October 30 due overdue Source
October 2026 · 3 updates
- 2026-10-22 (~4:32 PM ET, The Verge — theverge.com/tech): a second removal pass on the scraping language — Meta tweaked its business terms again to remove language suggesting it could freely train AI models on news sites — plus the first expert-grade evaluation of the reversal: Cornell's James Grimmelmann, to Nieman Lab: "This does nothing to restrict the 'associated with your products or services' [language] that would let Meta scrape entire sites." The ledgered October 30 effective date still stands. Watch: the final operative terms text at the effective date, a named publisher test case, regulator/AG reaction.
- 2026-10-17 (~11:04 AM ET, The Verge): the thread reverses — Meta has updated its commercial terms again to remove the scraping language; the "Meta may use any content from such domains, web pages, or URLs to provide, develop, improve, research, train" sentence no longer appears on the commercial terms page, following backlash (theverge.com/tech). One day after the terms were reported (October 16, Nieman Lab pickup) and thirteen days before the ledgered October 30 effective date. Watch: what replaced the language on the terms page, whether scraping permission reappears in another form before October 30, publisher/AG reaction.
- 2026-10-16 (The Verge citing Nieman Lab, ~10:02 AM ET): applies to business accounts promoting content across Facebook, Instagram, WhatsApp, Messenger. Watch: publisher reaction + opt-out mechanics before Oct 30, how "associated" resolves, a named publisher test case, regulator/AG attention. theverge.com/tech
Background
- The thread: Meta's revised business terms — effective October 30 — reportedly permit crawling and scraping websites "associated" with businesses on its platforms (publishers included), to develop/improve/train its AI products. Consent by fine print rather than lawsuit or license; the copyright-litigation and AI-search-traffic threads gain a Meta-terms datapoint. The effective-date claim is ledgered (Meta entry).