But like many kids who, regulators have found, commonly lie about their age to access social media platforms, she didn’t want to wait another year to join her friends on the messaging app. Hiding her age, she created an account that listed her as over 18 years old.
Discord has a lot of problems to resolve with reguards to community safety, but when it comes to authenticating another person's identity they'll need accurate information (even if it's only your age). It honestly did not help that a kid lied about their age just to get on the platform.
How could the site positively confirm a person's age without someone surrendering their government ID?
But such a thing is already being done with third party identity proofing companies. While Discord proper wouldn't get their hands on a users PII in your example, it would still create a real user ID attribution (something privacy advocates oppose).
The BC provincial government setup a system similar to this. Every resident of the province has (or has the option to get) a BC Services Card. This card is used for accessing government services and health care. This is a picture ID that includes your birthdate. Using this card, you can enable a BC Services account online (via the web or an Android/iOS app). Then, 3rd-party sites/businesses can use that for authentication / age verification / similar services.
The 3rd-party service never sees your actual information. They make a request for "is user over 18" or "verify this user is authorised" or similar. The request pops up on the app similar to MFA. You use your phone lock system to authorise the request. The 3rd-party service gets a token that says your identity is verified and to authorise access.
The Canadian federal government is working on something similar for accessing Revenue Canada and other federal services.
Similar to how banking online works via Google Pay/Apple Pay/Samsung Pay/Interac/etc where tokens are passed around rather than the retailers seeing your actual credit/debit card/bank account numbers.
Good luck getting a social media company to use these services, though, as that would require per-province/per-state/per-country/per-whatever setups. That would take too much effort compared to a hacked-together, homegrown, in-house "solution" that doesn't actually solve anything.![]()
Most DMVs are still struggling with email and online renewals. There's no such thing as "easy to implement" for a DMV. They'd have to farm it out to some private contractor.
I'm not sure I really want the DMV (or other government agency) finding out which sites I'm visiting, though. I also feel like this could easily become an oracle for identity scammers to verify information. You could find out which names were valid in a given jurisdiction, and by repeatedly manipulating the age field you could determine someone's birthdate.
If the only thing you have is a single boolean value from the verifier, that isn't a strong enough assurance if I were running the site. It's not enough to validate their age once, but continuously in anticipation of any changes that need to be made (age gates being raised for example), as well as possible erroneous or fraudulent validations as well.
I'm for collecting the bare minimum neccessary for this to happen, but what that means is determined by Legal.
You'd need to know the exact age because rules and regulations are going to differ. Laws and regulations could concern persons of various age groups.
On top of that, for your scheme to work the verifier would've had to put a reference in whatever message they're sending that could point to a user so it could positively prove their age via an audit trail. Hell, I don't even see the whole point of engineering this whole thing given that present day solutions already exist for this, and have been practiced everywhere that sells alcohol in America.
How do I get the token, then, and what prevents someone else from creating one on my behalf?
If the DMV isn't keeping track of what it signed then how can it be relied upon for validating someone's identity in this verification scheme?
If the authorities did an audit and believed that someone was underage on a service they weren't supposed to use, they would need to be able to follow that trail to positively confirm their age.
I mentioned the ID at the bar example because positively confirming someone's identity is the only way to positively confirm someone's age. A third party verifier could hand out "bar passes" to those of age so they wouldn't hand out their true ID to the barkeep, but the verifying entity would still need to retain the subject's actual ID information that the bar pass could point to, and the bar pass would contain information that could point to the persons true ID.
I think the problem is if you're accused of allowing a specific person that's underage, you need to be able to show that specific person gave you proof that you can trace back to their record at the DMV. An anonymous system doesn't provide that ability to prove what happened later.
Note that in most states, if you accept a fake ID you're still at fault for letting the person in, even if the fake is convincing.
That is only secure if the site and the DMV don’t both save the token and the user has no way of verifying that. The site would want to save the token to prove that an age verification was done. The police could demand a user’s verification token from the site and then demand that the DMV look up who authenticated for that token. The DMV could also sell the token associations to advertisers to fund the service like the post office does with advertising.