Facebook could be ~100% accurate at separating real and fake accounts if they exchanged data with the government to make sure that every user had a passport / social security number / other equivalent associated with their account. Is that the world we want to live in?
Some countries already do that - China, obviously. UK wants you to upload your ID when visiting a porn site (think of the children!). Japan has ID requirements for certain types of websites.
And of course, both Google / Youtube and Facebook have had a real name policy in place for a long time, to much objections. I've gotten a popup from Facebook asking me if the name of a friend who intentionally uses a fake name on Facebook to try and avoid her stalker is correct.
That would only be 100% accurate if the government approved ID was perfectly secure, and no one ever leaked databases of those personal details. Tens of millions of passport and social security numbers are already out in the wild.
How to make it secure? Estonian and Scandinavian countries have chip on them that a computer can read, but it's for doing gov and banking stuff. You can always fake a photo, they don't have much to compare it against to. At some point FB started to verify profiles actively, I didn't use my own name or took part of any political/marketing mumbo jumbo, but I quickly photoshopped an id with the fake name using a sample ID as a blueprint and the human moderator left me alone.
There is no "the government". Many countries don't have a central population registry, or at least not an accurate one. And refugees often have no valid identifying documents.
That's not the point and it's not even related to the discussion. I'm not sure why you are trying to bring in Big Brother nonsense.
The fact of the matter is that there are only 7 billion people on earth, and Facebook had a whopping 3 billion fake accounts. So how did they let it hit 0.5, 1,2 and 3 billion fakes before doing something? Surely with all their algorithms and AI/ML/Data Scientists they could have intervened sooner, but they didn't.
This is only for the last six months. They aren't letting the issue fester, but rather there are enough bad actors out there to try 3 billion times to create fake accounts. They are probably deleting 1000s of accounts per day that they detect to be fake
