Comment by aniviacat
10 days ago
Does anyone know how this is implemented?
If the proof can not be traced back to your identity, then what stops a person from creating large amounts of proofs and distributing them?
If the proof can be traced back to your identity, then... that would suck.
They use attribute based attestation which should be mostly anonymous. The long term goal was also to implement zero knowledge proofs which would make things like age verification fully anonymous, but because of technical reasons and development constraints that idea seems to have been postponed.
The reason you can't distribute a huge amount of proofs is that the app won't let you. To make sure the app won't let you, the app tries to verify that you're not running a modified app or a modified system environment. That's the remote attestation that "bans any android system not licensed by Google".
These tokens are signed and only usable for a limited amount of time so you can't just generate a million of them and sell them for others to use.
If the app can't rely on the system working as it should, it'll need to contain less privacy-friendly measures for limiting large scale token abuse.
For the proof to be traced back to your identity, you'd need to be tracked consistently across websites, possibly with the aid of the government itself. If ZKPs make it into the app, tracking you is basically impossible.
Of course, if you're authenticating with your full name and birth date, when opening a bank account for instance, you're not going to get the anonimity benefits. Still, you do get to see what party you've authenticated with and get a button in the app to request deletion or report suspicious behaviour if you think it was a scam.
The technical specification can be found here[1], with further details here[2].
Well, it's more like a framework, so not a ton of details. I've just glossed over it, but from what I can gather they have thought about it:
No personal data, especially no information from personal identification documents such as national ID card, is stored within an [Age Verification App Instance]. Only the Proof of Age attestation, specifically indicating "older than 18", is utilized for age verification purposes
Stored Verification(8b): [Relying Parties] may optionally store information derived from the Proof of Age attestation in the User's account, allowing the User to bypass repeated verification for future visits or purchases, streamlining the User experience. In this case, authentication methods such as WebAuthN should be utilised to ensure secure access while enabling the User to choose a pseudonym, preserving privacy. Risks in case of the device sharing should be considered.
[1]: https://ageverification.dev/Technical%20Specification/archit...
[2]: https://ageverification.dev/Technical%20Specification/annexe...
Even if they can't be traced back to a name/photo identity, it would still be a privacy disaster if you could only make one proof per service.
If a user can only make one then they'll have to use that identity with that service forever. That's a nightmare for privacy. Sometimes people need another account, unknown to their employer/family/friends. People should be able to make multiple accounts without those being tied together through a common "age check" identifier. But, of course, there is no way to prevent those from being distributed.
At some level I believe that's the purpose behind some of this. If someone can only have one proof, then someone can only have one account to speak with. They'll be easier to monitor, easier to identify, easier to silence. That's why I think these types of laws and behaviors should be resisted and protested.
I've mentioned in a previous comment that it's telling that big tech isn't resisting these totally-just-coincidental ID laws coming from western countries. It supercharges their surveillance and tracking abilities, and widens their moats.
Also, porn is a smokescreen. The definition of "adult" content will rapidly expand, and these put the ID issuers in censorious a position of control over people and services. Nothing stops a government attestation server from rejecting a request because someone is blacklisted from "mass communication services" because they're a felon, protestor, LGBT activist, etc... or because a service has fallen out of favor.
The idea is, that you have a 'digital ID' on your phone, tied to your real identity, that will today be used to prove you're 18, but when the infrastructure exists, it will be used for other stuff too... like needing to attach your real name to any social media account (you already have an app that does that on your phone for the 18+ thing, so adding real name is easy to implement), and that will greatly affect freedom of speech.
This is the pr on it [0]. It was linked on hn at the time too [1]
For all the shit Google deservedly gets they seem to be genuinely trying to implement good and privacy preserving solutions to a lot of these problems.
The issue of course is that there's essentially no way to do all this stuff with software and hardware the user actually controls themselves, so you end up with hard requirements that you use big tech as gatekeepers.
This is the slippery slope that IMO eventually ends the open web.
If you take that outcome as inevitable, which at this point I basically do given all the forces lined up to restrict access to information, I suppose Google is about the best steward you could hope for.
[0] https://news.ycombinator.com/item?id=43863672
> If you take that outcome as inevitable,
I don't and I wish Google et al would take a god damned stand against it. All it takes is 2 or 3 big companies to just not play along with the destruction of the open internet (the very same responsible for their genesis and incredible success), and the bureaucrats will eventually relent. Unfortunately they've chosen the path of least resistance, which also is the path of regulatory capture to their sole benefit. Sad to see that win over the ideals of the early net.
I agree in principle but as time goes on I have found that the free and open internet as we know it already no longer exists in practise. Theres like 5 places to go on the internet these days - your social media platform of choice, your short form content platform of choice, youtube, perhaps an AI platform, and 1 misc place of your preference. And this loop of crap seems to demand more and more of your life.
I went on youtube in bed last night to watch a 10 minute video (that I knew I had to search for to find - it was a specific one), but the app opens to shorts and they're so damn stimulating that it was 30 minutes before I finally got to the vid I wanted. I started with pure agency and was immediately thrown off course. Say what you will about my discipline or habits, but imagine the affect this has on less... aware individuals such as children.
Walking around the world you see everyone buried in their phones.
There are aspects of this initiative that I totally welcome, if it has the result of some level of de-interneting. The argument is always "they do it to protect children first, then it comes for everyone". I hope they increase resistance for the end user. I agree its sad, but what we have currently is truly awful, and less of it is a good thing.
I understand that it may not have that effect and end up in the "worst of both worlds" situation. But I don't wan't google fighting any battles for me anymore. They might try on occasion to be respectful but their bottom line is to own my attention.