Meta being Meta. No surprise there.
News are posted elsewhere.
You authorize us […] to […] create test credentials to […] crawl the domains […] associated with your products or services.
Meta may use any content […] to […] train […] Meta Company Products.
Basically, in exchange for being allowed to continue posting content on Facebook, you’re giving them permission from now on to aggressively scrape your own sites for content to train their models.
To quote a famous Sith Lord: “I am altering the deal. Pray I do not alter it further.”
I mean, this is fine. The “content” most businesses publish where meta can see it is all sanitized advertising nonsense anyway. This will create for meta the least useful ad generator; assuming that such a thing could actually be useful.
To be clear, this is not restricted to content posted on Facebook. It’s any content posted on any website owned by the customer.
And specifically gives Meta the right to create “test credentials” on sites and scrape resultant deep content such as product docs that are only available with an account. This is far more invasive and concerning than traditional site scraping and indexing.(refs: https://www.theregister.com/ai-and-ml/2026/10/16/meta-grants-itself-right-to-scrape-content-from-business-customers-for-model-training/5303658 , https://m.facebook.com/legal/commercial_terms_preview )