Meta Halts Terms Update That Would Let AI Train on Linked Sites
Meta has paused a commercial terms change that would have forced business account holders to let it crawl and train AI on outside websites linked from their Facebook pages. The update was pushed from Oct. 30 to Nov. 20, then shelved outright after court injunctions in Brazil and Norway, a U.K. antitrust probe and pushback from two U.S. senators.
- Original source: tokenpost.com.
- The source published it on Nov 8, 2026 at 5:19 PM.
- No second outlet has reported it yet. Treat it as a single-source claim.
Meta has stopped the clock on a commercial terms overhaul that would have tied continued business use of a Facebook page to permission for AI training on linked outside sites. The company says its existing terms stay in place while it keeps talking to stakeholders. No replacement date has been announced, leaving Nov. 20 — the second deadline it had set — as a date that will not be met.
The proposal reached well past Facebook itself. It covered material on external websites reached through pages run by business accounts, and it made crawling those sites a condition of commercial use, with no separate opt-out. Meta would have been free to use whatever those crawlers collected to build and train its products. Publishers were also asked to hand over test credentials so Meta's crawlers could get behind paywalls.
Resistance built fast. Preliminary injunctions in Brazil and Norway blocked the terms after organizations took them to court, and the U.K.'s Competition and Markets Authority opened an investigation into whether Meta was leaning on its market position to extract publisher material for AI training. Sens. Adam Schiff, D-Calif., and Marsha Blackburn, R-Tenn., wrote to CEO Mark Zuckerberg asking him to withdraw the update outright.
Meta tried a rewrite first, stripping an explicit mention of training while keeping language about developing its products. Theo Bamber, chief executive of the U.K.'s News Media Association, said that wording still worried publishers because collected content could be used to develop business products. Meta countered that its crawling practices were unchanged and that site operators could use robots.txt to signal their preferences — though the planned terms never promised to honor those signals.
On Oct. 28, Meta Vice President of Communications Andy Stone posted that consultations with industry groups and leaders could run past Nov. 20, and that Meta would delay implementation if they did. That made the November date conditional from the start. It is now off the table entirely, with the terms paused and no new timeline on the books.
TopicsMeta · Mark Zuckerberg · Andy Stone · Adam Schiff · Marsha Blackburn · Theo Bamber · News Media Association · Competition and Markets Authority
Source: tokenpost.com
Written by Paparazzi with AI. Not financial advice.
More stories
- Global & Socialindependent.co.ukCaught by @adiwijaya0viewsDeer struck by SUV flies into oncoming Jeep, killing Missouri driver
- Global & Socialbbc.co.ukCaught by @adiwijaya0viewsThirteen migrants die as boat sinks off Crete, 18 pulled from the sea
- Global & Socialqz.comCaught by @hafizutomo2120viewsThree charged with AI impersonation plot that stole $129 million
- Global & Socialcbsnews.com0views14 dead off Crete as Greece pulls 180 more migrants from the sea

