Meta is changing the terms for its business tools on October 30. Three legal experts told Andrew Deck and Hanaa' Tameez at Nieman Lab that the new wording gives Meta broad permission to scrape the websites linked from your business accounts and use what it finds to train AI. Meta disputes that anything changes in how it crawls publishers' sites.
Nieman Lab published that on October 15. If your business has a Facebook or Instagram presence, you have two weeks to work out who's right. Don't treat this one as another terms notice to click past.
The updated terms, in the clause Nieman Lab reproduces, cover websites associated with your products or services. They allow Meta to collect content from those sites automatically, including to train Meta products. You don't sign anything. If you keep using the business services, you've accepted.
That's a much bigger deal than the old worry about what happens to a photo you post on Facebook. Your website might carry articles you commissioned, photos you license and work you got under separate agreements, and your rights over each can differ. Putting a link in your business profile doesn't make those differences go away.
So the person who runs your Instagram account shouldn't be the one deciding this. They may have no authority over licensing for the whole publication. Bring in whoever is responsible for your website's content rights.
Meta says robots.txt still protects you
Meta told Nieman Lab the update just consolidates existing language, and that site owners can still control access with robots.txt, the small text file that tells crawlers what to skip. The legal experts weren't convinced. Their question is what happens when a technical instruction says one thing and a contract says another. Nobody knows yet whether a court would back the broadest reading.
Think of robots.txt as a sign on your door. The contract is a permission you've agreed to. When a platform says the sign still counts, you want that confirmed for your own sites, not assumed.
Before you decide whether to keep the accounts, find out exactly which sites and which content the terms would cover. That's a specific question your account manager can take to whoever handles your company's contracts.
Meanwhile, Google is paying publishers
The same day Nieman Lab's story ran, Google said more than 400 publications are now in its News AI pilot. "We pay participating news publications for the delivery and use of enhanced content," Google writes. That content feeds features like article overviews and audio briefings, and the new partners include local news organizations as well as large publications.
It's a separate program covering different services, so it settles nothing about Meta. It does show that paying for news content and negotiating a license are still on the table. Keep that in mind when Meta's version is a permission bundled into your social-media tools.
What to do before October 30
Save the notice and the terms as they stand. List every account your business uses and the sites each one points to, then ask whoever manages your content contracts to review the new permission. Where it's unclear how the terms interact with robots.txt, ask Meta directly, and keep a record of what it tells you.
Meta is rolling out more models for businesses at the same time, including Muse Spark Open. The next ones will be trained on something. By October 30, you should know whether that includes your website.
Cover: AI-generated editorial illustration by TMRW using Meta branding. The document and newspaper are illustrations.
