The only entity able to connect you in this case is the identity verification third party. The premise is that a government-backed identification system is more secure than a rando private company.
Private company asks government “hey is this user real and unique”, government replies “yes”. Private webiste does not need to know your ID. No identifying element needs to be transmitted by the government.
Of course some private companies will need more, and in that case the user, you, can grant them access to data, much like the current authentication systems using Google accounts & co.
In which case the flow would be:
- Rando insecure company asks government "is this user real and unique? I need their name"
- Government website asks you "this rando company wants to know your name"
- You accept
- Goverbment replies to rando insecure conpany “yes, user real, name is X”
That’s how it should be.
offspec@lemmy.world 2 days ago
??? This is just textbook sso/openid but backed by the government. There’s nothing intrinsically insecure about having third parties send you directly to a trusted government site for authorization.
fluffykittycat@slrpnk.net 2 days ago
And how does that prevent hacks that reveal the connection between ID and user account ?
offspec@lemmy.world 2 days ago
What connection do you think a third party is saving when using openid? Generally speaking the only thing the third party needs is your identifier which in most cases is just an email. It’s no more devastating for the user base for that information to be leaked than it is when they’re handling authorization themselves. I personally think using a government backed authorization platform is a terrible idea and something completely liable to be abused by those in power, but it would objectively be better than trying to have every single service store your personally identifiable information themselves.