again, even if I were the one doing that Google search, with the domain examples you provided, I wouldn't trust one of them.
Like, common sense on the interwebs just continues to disappear. Gullibility seems to have increased as critical thinking and coming to logical conclusions are disappearing.
Yes, that is the point of the TFA, but I was commenting on what data was posted online well before the AI was "born". I'm guessing that Meta can fix it with a prompt that tells the system it is not allowed to verify FB phone numbers or there are NO phone numbers for the public to use to contact FB but do not inform the user that there are not phone numbers. Only deny any phone number is a valid phone number for FB (or Meta).
All you'll find is results like https://gethuman.com/phone-number/Facebook or https://facebook-pay.pissedconsumer.com/customer-service.htm... or the Quora post that all have numbers that almost certainly goes to a scammer.
Bonus points when it makes it into the AI models (as happened here) where they repeat it verbatim as if it were the truth.