The original idea for the ID verification was broken by design anyway. The only safe and secure way is a chain/tree of trust, e.g. with PKI, where you could generate some certificate just for that particular service, while keeping your root key safe. Then, in the case of leak, the most you lose, is one particular key for one particular service that could be immediately revoked. You could even slap zero-knowledge proofs for particular properties (e.g. if the person has a driver license or not) without de-anonymizing the account. In the rare even of root key leak you should be able to physically go to the authority and make a new one, while revoking the old key. I don't see any other better alternatives than this.
This is effectively the EU age verification system.
Your government (which already has all your details) generates certificates and you just give those out. The other side can the use simple public/private key verification to ensure the cert is valid. Also government does not get information who you gave the cert to and if you create a bunch and single use them the other side can’t follow you between uses using the certs.
The EU age verification system requires tying yourself to Google/Apple [0] (ie will not work with GrapheneOS) which is a non-starter. This means you will need a mandatory Google/Apple account. What if your Google account gets banned?
eID PKIs have very little in common with the web PKI. There's a national root of trust with strong attestation. It's a very simple trust relationship. You already trust the respective government to issue IDs.
Plenty of European countries have an eID CAs and it works fine. The PKI part is a solved problem.
Doesn't even need ZKP, the CA can just issue an attestation.
I've experienced this "solved problem" when visiting Germany during COVID. On every entrance to a mall there was somebody with a scanner device, and they only let you in if the scanner showed a green mark. I've been fully vaccinated (not EU) but my code didn't show a green mark on their scanner and I was promptly denied entry. The solution was to show them my German friend's code on my phone, this registered just fine, so I could enter anywhere.
So the whole "eID/CAs/ZKP/PKI" mumbo jumbo can be easily fooled by a gif file.
That part seems to qualify as an unsolved problem. But could anyone have taken the scanned data (or the GIF file) and used it to open a bank account in your friend's name? That seems like the main issue that is genuinely solved by correct implementation of this type of system.
Already today nobody can open a bank account in my name with just a picture of my passport, as the original would be required. My passport doesn't have any of the "eID/CAs/ZKP/PKI", so the question of "what exactly the addition of it solves" remains open.
My national ID card supposedly has some of it, the 17-year olds who want to pass as 18-year olds usually show a doctored gif file of their ID card, with a year of birth one or two years before the actual one; this works in ~98% of the cases.
The three times I've needed to provide a scan of my passport were: to enroll in a university course, to buy from an e-commerce site, and to become an app developer. None of those orgs really needed a scan of my passport, which can't be revoked like a cracked password, and will now sit unencrypted somewhere until the end of time or until they are hacked and subsequently shamed into handling their customers data more like radioactive waste.
The US Government is one of the reference implementations of PKI.
Unfortunately, IDs are issued 50 different ways by the less competent states.
Combine that with accusations that getting new IDs constitutes systematic racism (a widely held belief on HN), ignoring that the ruralest of India has been able to do this successfully, and you're not getting digital ID any time soon.
The difficulty for the US is people seem to be against a Federal Government ID. India doesn't seem to have this stigma and hence rural India can solve this problem.
I don't think people are against it, we already have the social security identifiers as a government layer... it's just that no one in the government is willing to do it for free in a way that is accessible to everyone
IDs being required for voting is essentially a poll tax the way that it's argued for in the US. There's a minimum amount you have to spend to even get one that will expire in a certain amount of years.
That's also assuming you can get all the documents you need for the initial ID. If you don't already have all the essential documents you'll need multiple appointments at government facilities. The local social security office usually has just one location. Mine required time off from work because they were open from 10 till 4 four days a week. This was just part of the process to get an ID for a child when we only had a copy of a couple documents instead of the originals.
Getting an ID for my grandma who made the mistake of being born black in the rural South during Jim Crow was a year long process, since she didn't even have originals of most of the documents. It took that long and was handled by one of her children that was a lawyer. For her we were getting an ID so the family could take her traveling, years before it was turned into a political issue to disenfranchise people.
Nobody arguing for the ID laws ever argues for raising their taxes for an ID that all citizens are guaranteed with the supporting infrastructure. They're arguing for it because they know people they don't want to vote will have trouble getting one
I think with passkey you don't own the private key. It's in your device and managed by the OS. That's one of the reasons I don't use passkeys (the other being that if I lose the device I can't access my account)
> think with passkey you don't own the private key. It's in your device and managed by the OS. That's one of the reasons I don't use passkeys (the other being that if I lose the device I can't access my account)
It is not true. You can move passkeys between OSs if you have a password manager or an OS that has this ability. For example, I store my Passkeys in iCloud Keychain and I have them synced on all my Apple devices.
There are cases though, where the website can force the requirement of a device bound passkey, but that is something it is not very likely you will encounter.
You can, with for example KeePassXC. But following passkey development it does seem that in time they will want Google or Apple to own your private key and for you to use it through them.
Yes, you can. Either you can physically own the key on the physical device (i.e. Yubikey/Google Titan/HSM) or you can use software (KeepassXC, Bitwarden).
Most operating systems and browsers come with a sync mechanism that many people default to, but it's no more than that: the default.
As for account recovery, most websites have a way to recover your account when you lose your password, there's no reason why that wouldn't work for passkeys. Every website with passkey access I've used so far makes passkeys optional and forces you to set a password already. If they switch their default to passkeys and add a password as an optional step, nothing would change.
My passkey is synched with iCloud which has all hardened settings set to on, so if I switch devices (I have) its seamlessly. You can also export Passkeys is my understanding but I have not tried it.
When I'm on non-Apple I can scan a QR code on my phone and it will validate it the same way, using bluetooth to send enough info to the target computer to get me logged on.
Bluetooth is actually only used to establish proximity. Both your phone and the browser connect to an internet relay, which lets the phone essentially pretend to be a USB security key connected to the computer.
My passkeys are in my self-hosted Vaultwarden. If I can't put them there, I don't use them.
It's not all sunshine and roses, though. Despite having Bitwarden set as the only passkey provider in my Android setup, the phone persistently only offers me Google. Which is empty, because as I said, I won't use one tied to things I can't control. Works great on desktops, though.
Passkeys can theoretically require you to be on hardware, I haven't found anything yet that requires that.
I don´t think I've ever come across a service that only used passkeys. Username/email + password + 2FA is usually the primary form of verification. There's usually a way to recover your account through email.
I use passkeys only, for the backend admin app for my app.
They work fairly well. I use a login ID/PW to set up the passkey, then the passkey, for everything else.
Passkeys work well, but I feel as if they are still a bit too “fiddly” for your average mensch.
I’m pretty sure the private key is transferable, but Apple keeps them in the Keychain, so they seem locked into the OS. On my machine, I can also keep passkeys in 1Password.
This is the answer. The more failures, the more justification for more draconian restrictions of civil liberties.
I can hear the defense now: "Oh, yeah, you blame the honest, good, handsome people trying their best to protect you and you let the hackers off scot-free! We must make sure that hackers don't have access to the tools that aid them to commit these crimes, like books and computers. Anyone could be a hacker."
That is an unfair conclusion. These people run complex networks like the rest of us, they probably have a range of detection systems and, also like the rest of us, an almost impossibly large attack surface to consider internally and on their supply chain.
The problem is that it is really, really hard to make something secure even if you try and follow all the best-practices you know.
I guess the awkward bit is marketing everything as certificate this, accreditation that and overselling how secure it is although I don't really know how else you would word it, "as secure as we know how"?
Funny was just testing the pilot of the Irish Government Digital Wallet. Definitely seems like the way forward if we're intent on doing identity verification. I'd rather the government mediate this than a bunch of random 3rd parties.
We also have a Danish wallet now, AltID, which implements an anonymized (assuming no collusion between issuer and eavesdropper or service provider) age verification protocol based on batches of single-use tokens which contain no personal information (except that they can be traced back to you by the issuer).
It's been released and in production since summer. Since then, several social networks have apparently started A/B testing age verification for their EU users, but how many of them actually integrate with the anonymous solution that is now available and in production? To my knowledge: 0. They all use Persona.
This highlights one of my main criticisms of EU's naive approach to regulation of tech companies. They fail to realize that any regulation that they impose will be complied with in the most malicious way possible, which is how we got cookie banners with dark patterns instead of a simple HTTP header saying no thanks to cookies.
Well, that's at least a positive. I hope this means they are also forced to accept the weakest credential type, and that they can't require you to present the identity proof that the wallet also provides when verifying your age. Because if they are allowed to do that, then that is what will happen.
In Finland they outsource the system to banks and telephone operators. Its a very strange system. As far as I know, its not possible to access government services just by being a citizen. You also have to have an account with one of these third parties to get in.
You can, but there's a couple extra hoops to jump through. You first have to go to the police station and get a state ID because banks don't accept US passports or Finnish residence permits for whatever reason. Then you can open an account at Nordea. Then you can go through Nordea's system for e-identification.
probably gov will outsource to 3rd party for gov to build system to track and manage ID. Sometimes though also to manage, as in Denmark's MitID mainly managed by NETS under government set rules.
As a citizen under the France Passoire[1] and in an increasingly fascist chauvinist nationalist drifting in the geopolitical landscape, I wouldn’t be that found of delegating too much of these responsibilities to some centralized governmental institutions.
Note that’s this is not here some rant against any governmental power, just that in context, large private group use them as puppets and shrink their budget which of course impact quality of deliveries (not shaming the people who do the hard job without the relevant means). And while more distributed governmental topologies would have their own caveats, at least it would less likely offer opportunities for single point of failure.
On the end of the day, it is the state that issues these ID documents. So if you let the government go bad, IMHO the form of the documents does not matter that much.
During the totalitarian communist rule in Czechoslovakia, the state would regularly interfere with passports of people considered not loyal enough - withholding them outright or inventing extra paperwork that was necessary for the border police to let you out of the country. They also controlled all supply of foreign currency, both in an out.
Then if someone was actually allowed to travel outside the country but failed to return, their family and relatives would be punished, including demotion at work & prohibition of higher education.
So if your government goes bad, this is what will happen - the form of the ID takes at that point does not make much difference.
I live in the UK and was having this exact discussion with someone recently - I'd actually prefer Apple to be the owners of my digital identity over the UK government who would happily throw you in jail for expressing support for Palestine Action.
I think this is a bit of a non sequitur. The government will still be able to throw you in prison in either scenario. They will also be able to compel disclosure of your records in either scenario. So the difference is really just that in the scenario where apple verifies your identity there's an additional actor in the mix over who you have zero influence. It seems strictly worse.
Your digital identity would then be under the jurisdiction of the US government, who (despite talking a big game) are arguably worse for that same sort of behaviour.
If anyone has to have this type of control, better it be a local national government that you can in at least some small way influence.
I understand the sentiment, but your government you can actually fix by voting.
If Apple (or another large international company) suffers from decreasing margins, gets a new CEO and decides to turn the data it sits on into money there is absolutely nothing you can do, and you might in fact still stay a "forced" customer because of network effects (=> just consider whatsapp being an important communication channel in many places worldwide).
I think this attitude in general is often harmful; if your government sucks, fix the government instead of making yourself dependent on some quasi-monopolist private company.
> You can fix companies by voting a government too
Indirectly only. This is typically also always too late; instead of doing "the right thing" in the first place, companies are disincentivized by regulation from doing "bad things" again.
Regulations are like scar tissue, they don't help against getting burnt in the first place.
Preemptive regulation typically sucks, and is admittedly extremely difficult to get right; most governments don't even bother trying.
Corporation know this and exploit it ruthlessly-- there are almost never consequences as long as they keep to the letter of the law, even when acting with intent, against better knowledge and causing astronomical damage to society (just consider the whole leaded gas disaster for an extremely clear example).
> throw you in jail for expressing support for Palestine Action.
For those unfamiliar, you are, of course, allowed to peacefully protest in support of Palestine in the UK.
Palestine Action is a specific group that was controversially labelled as a terrorist group after they broke onto a runway and spray painted military planes.
It would be best for any such ID system to capture only the minimum data to fulfil their function. Pre-WW2 Dutch records also listed religion, which made it very easy for the Nazis to round up and murder them. Some examples from Kansas show that even recording sex/gender is risky - state law requires driving licenses to record the original sex/gender from the first birth certificate leading to trans women having an M sex/gender marker getting arrested and prosecuted for identity fraud.
LOL, no, not in this side thanks. Be it by giant transnational corp or centralized all powerful government, the ruin of the noosphere is equally awful.
You know, it always bugged me that the NSA (and more directly Google, and my phone company, and so on) know where I was at exactly this time a year ago, but that I do not.
If you are in California the DMV makes tens of millions of dollars a year selling all the data you give to the DMV, which is why I give them a P.O. Box.
CADMV claims on their web site that they cannot accept a P.O. box as a residence address. I have yet to find anything in California state law supporting this policy, though IANAL. Their enforcement seems to be quite lax.
It is a REAL ID requirement. It is federal law. States issuing REAL ID compliant identity documents must mail them to your physical address. USPS provides that data for address validation.
DPVCMRA = delivery point is a commercial mail receiving agent. Any sort of location with PO Boxes.
As someone who spent years homeless off and on, I resent REAL ID discriminating against the unhoused, so I choose to use my privilege to reject it. I do not have or need REAL ID and my passport does not need a current residential address.
I used to work for my state's DMV. They'd sell vehicle registration data to various companies, which is how and why you get those "we've been trying to reach you about your car warranty" phone calls. I don't know about CA, but KY had a problem with tracking who ordered and who paid for that data. When I worked there, we found a number of "purchasers" who only paid for Year 1 but stopped paying afterwards.
Federal law requires state DMVs to supply that data to the car manufacturers. So if you own a Chevy, they have to send your data to Chevy in case there is a recall.
I suppose one other industry interested in this information is the bounty hunter industry (bail enforcement). I believe there are several platforms for licensed bounty hunters where they can receive this and similar information about a person/vehicle.
That's why I say "Our lobbyists have more money than your lobbyists". Every state has sunshine laws to show who the lobbyists are, what they lobbied on, and to whom. Some states separate those lobbyists into legislative & executive branch lobbying.
I suggest you look at who voted for those bills, who lobbied them and who hired those lobbyists.
How exactly does that work? How can you sneak a live feed past detection systems? It is incomprehensible to me, considering this is highly regulated and sensitive data. It is just open ports sending what they shouldn't be sending all the way out or what?
Brian Krebs' article makes a good case for the ID source being a harvester on the internal Hertz Car Rental network, and likely other similar consumer services that log ID for asset security and recovery.
These are hardly military grade networks, as long as the driver licence scans make it to the database and can be used to identify and recover damages from accident or theft it's unlikely anybody has cared much past that functionality.
I’ve been following the development of the drivers license sharing system from Apple where different fields can be selected; are there any implementations of PKI based identification systems where multiple certificates can be generated and revoked when compromised?
I’ve often thought that replacing the US social security number with a more robust root key makes for a fun thought experiment. Hard to imagine how such a system could securely serve so many people but passports with embedded chips seem to be doing okay.
Not in the US. Several EU countries have PKI systems integrated with identity documents that let the requester to ask for age (for example) and then only age is supplied. But their PKI systems are for the whole ID document.
As for the passport, the key/PIN you need to authenticate to the chip are printed on the page with the photo. Otherwise "hackers" can only determine nationality of passport. The standard is ICAO 9303.
SSN was never intended for identification. My original card, issued in the 1970s was clearly marked "not for identification". In the original numbering system, the first 3 digits identified the office/area where the card/number was issued and the next 2 digits identified the filing cabinet. 700s were set aside for railroad workers (until 1963) because the legislators did not want railroad workers to be included in social security.
No worries! Governments who used this company are taking responsibility and now have a plan to, at the very least, replace all IDs they forced people to expose and to make sure the old ones are unusable!
That's a sarcastic joke. It's how governments demand private companies react, but ...
> This week a massive new data breach has been revealed that should put the nail in the coffin for the idea that any sort of age or identity verification could be safe.
Yeah just like how the multiple breaches and utter negligence from the incumbent credit bureaus killed the credit file managed by private companies.
I think there's a number of people reading this who clearly didn't detect the satirical nature of this single sentence. It's blunt and obvious, but even so...
I still don't understand why the simplest approach isn't used: ban kids from using the Internet unsupervised. There's really no good reason why a six year old should have internet access.
The argument is that there are parents who are too stupid/lazy to enable parental controls on kids devices and society has a duty to protect kids even if their parents are negligent. Also, kids interact with other kids, so even if you do everything right your kids wind up with access/peer pressure through the kids with bad parents.
I dunno if I agree but I think that's the thrust of it.
This is precisely why the authority doing these checks needs to be the government that already issues the IDs.
Using ZKP as the EU proposes is the only way to prevent this data being leaked to unreliable third parties and leaves the knowledge with the institution it derives from in the first place.
I don't know why HN rails against it constantly, it is the obvious technical and organizational solution to this issue.
Government systems leak information all the time. The type of institution managing the data makes little difference. Its how the institution manages the data that matters.
In the olden days, the doc would hold your nuts and tell you to turn your head and cough. This was to detect if you had a hernia (which was a medical exception).
The bunghole inspection was to see if you were the receptive partner in anal intercourse because being a gay man was a criminal offense.
Allegedly, lesbianism was never outlawed because Queen Victoria was so offended by the idea of it that she prohibited the legislation from even being heard in Parliament.
so... for military grade identification systems instead of face photo with id document next to it, you make banghole photo with id document next to it?
Concrete ZKP age verification schemes are hardly zero knowledge.
Imagine your idealized ZK address verification scheme. It would go something like: I show up at a website, it sends me some challenge, I send back a signature of the challenge that could only be made by someone with an of-age ID, but without specifying who. Everyone is happy.
Now little Johnny borrows my ID, and uses it to setup some oracle that provides ID validation for every kid and bot in the country. Woops.
To stop that you must compromise the idealized zero knowledge properties of the scheme, and in doing so you create the potential for harm/risk for everyone.
Sure, it's better than sending an ID card live feed to the dark web, but the risks of ID card theft are at least somewhat easy to understand.
Some of the threats to human rights don't even require the departure from the 'idealized' model-- as even the idealized model requires an ID issuer to issue the of-age person an ID. And so if the ID ZKP is widely required then the issuer can unperson you by simply declining to issue you an ID.
The original idea for the ID verification was broken by design anyway. The only safe and secure way is a chain/tree of trust, e.g. with PKI, where you could generate some certificate just for that particular service, while keeping your root key safe. Then, in the case of leak, the most you lose, is one particular key for one particular service that could be immediately revoked. You could even slap zero-knowledge proofs for particular properties (e.g. if the person has a driver license or not) without de-anonymizing the account. In the rare even of root key leak you should be able to physically go to the authority and make a new one, while revoking the old key. I don't see any other better alternatives than this.
This is effectively the EU age verification system.
Your government (which already has all your details) generates certificates and you just give those out. The other side can the use simple public/private key verification to ensure the cert is valid. Also government does not get information who you gave the cert to and if you create a bunch and single use them the other side can’t follow you between uses using the certs.
The EU age verification system requires tying yourself to Google/Apple [0] (ie will not work with GrapheneOS) which is a non-starter. This means you will need a mandatory Google/Apple account. What if your Google account gets banned?
[0] https://github.com/eu-digital-identity-wallet/av-doc-technic...
I don’t really trust anyone to get PKI right. There’s enough mistakes in the www realm that pretty well prove bad actors will get through.
The alternative is do it offline.
eID PKIs have very little in common with the web PKI. There's a national root of trust with strong attestation. It's a very simple trust relationship. You already trust the respective government to issue IDs.
Plenty of European countries have an eID CAs and it works fine. The PKI part is a solved problem.
Doesn't even need ZKP, the CA can just issue an attestation.
You’re conflating the real life need of ID, with trust in the organisation.
Sorry. Wrong.
I've experienced this "solved problem" when visiting Germany during COVID. On every entrance to a mall there was somebody with a scanner device, and they only let you in if the scanner showed a green mark. I've been fully vaccinated (not EU) but my code didn't show a green mark on their scanner and I was promptly denied entry. The solution was to show them my German friend's code on my phone, this registered just fine, so I could enter anywhere.
So the whole "eID/CAs/ZKP/PKI" mumbo jumbo can be easily fooled by a gif file.
That part seems to qualify as an unsolved problem. But could anyone have taken the scanned data (or the GIF file) and used it to open a bank account in your friend's name? That seems like the main issue that is genuinely solved by correct implementation of this type of system.
Already today nobody can open a bank account in my name with just a picture of my passport, as the original would be required. My passport doesn't have any of the "eID/CAs/ZKP/PKI", so the question of "what exactly the addition of it solves" remains open.
My national ID card supposedly has some of it, the 17-year olds who want to pass as 18-year olds usually show a doctored gif file of their ID card, with a year of birth one or two years before the actual one; this works in ~98% of the cases.
yes, just as age verification can be fooled by an older friend who gets the check mark.
Reminds me of
(Bruce Schneier dug into origins of this here: https://www.schneier.com/blog/archives/2026/05/laurie-anders...)or just not ask for it at all.
The three times I've needed to provide a scan of my passport were: to enroll in a university course, to buy from an e-commerce site, and to become an app developer. None of those orgs really needed a scan of my passport, which can't be revoked like a cracked password, and will now sit unencrypted somewhere until the end of time or until they are hacked and subsequently shamed into handling their customers data more like radioactive waste.
The only times I had to provide a scan of my passport was to cross a border:
I have another ID for authentication inside my country, which by the way, is not a driving license (which is used only for car driving usage).The US Government is one of the reference implementations of PKI.
Unfortunately, IDs are issued 50 different ways by the less competent states.
Combine that with accusations that getting new IDs constitutes systematic racism (a widely held belief on HN), ignoring that the ruralest of India has been able to do this successfully, and you're not getting digital ID any time soon.
The difficulty for the US is people seem to be against a Federal Government ID. India doesn't seem to have this stigma and hence rural India can solve this problem.
I don't think people are against it, we already have the social security identifiers as a government layer... it's just that no one in the government is willing to do it for free in a way that is accessible to everyone
I think this is the horse shoe theory where the far left and right are against it. https://www.aclu.org/press-releases/aclu-conservative-and-li...
I know people who don't have a social security number because their parents didn't want them to be tracked by the system.
It is a problem for millions of Americans: https://voteriders.org/article/who-doesnt-have-an-id-in-amer...
Systemic racism is very much a thing, and while perhaps not foundational in this particular issue, we see still see political fuckery that definitely targets by race: https://www.nbcnews.com/politics/2026-election/tarrant-count...
IDs being required for voting is essentially a poll tax the way that it's argued for in the US. There's a minimum amount you have to spend to even get one that will expire in a certain amount of years.
That's also assuming you can get all the documents you need for the initial ID. If you don't already have all the essential documents you'll need multiple appointments at government facilities. The local social security office usually has just one location. Mine required time off from work because they were open from 10 till 4 four days a week. This was just part of the process to get an ID for a child when we only had a copy of a couple documents instead of the originals.
Getting an ID for my grandma who made the mistake of being born black in the rural South during Jim Crow was a year long process, since she didn't even have originals of most of the documents. It took that long and was handled by one of her children that was a lawyer. For her we were getting an ID so the family could take her traveling, years before it was turned into a political issue to disenfranchise people.
Nobody arguing for the ID laws ever argues for raising their taxes for an ID that all citizens are guaranteed with the supporting infrastructure. They're arguing for it because they know people they don't want to vote will have trouble getting one
So… requiring an ID to buy a gun is also an unconstitutional tax on a right?
While that can certainly be true it is still generally safer than username/password for authentication.
> the most you lose
Yep that's the only thing you lose, apart from a huge number of literal images of kids in the hands of literal criminals.
> I don't see any other better alternatives
Not doing age verification!
age verification is the current mandate of multiple governments, not doing it for many services means shutting down
I don't mean the services I mean the governments.
That sounds like this prediction from 2003
https://www.fourmilab.ch/documents/digital-imprimatur/#SI_an...
Passkey?
I think with passkey you don't own the private key. It's in your device and managed by the OS. That's one of the reasons I don't use passkeys (the other being that if I lose the device I can't access my account)
> think with passkey you don't own the private key. It's in your device and managed by the OS. That's one of the reasons I don't use passkeys (the other being that if I lose the device I can't access my account)
It is not true. You can move passkeys between OSs if you have a password manager or an OS that has this ability. For example, I store my Passkeys in iCloud Keychain and I have them synced on all my Apple devices.
There are cases though, where the website can force the requirement of a device bound passkey, but that is something it is not very likely you will encounter.
You can, with for example KeePassXC. But following passkey development it does seem that in time they will want Google or Apple to own your private key and for you to use it through them.
Yes, you can. Either you can physically own the key on the physical device (i.e. Yubikey/Google Titan/HSM) or you can use software (KeepassXC, Bitwarden).
Most operating systems and browsers come with a sync mechanism that many people default to, but it's no more than that: the default.
As for account recovery, most websites have a way to recover your account when you lose your password, there's no reason why that wouldn't work for passkeys. Every website with passkey access I've used so far makes passkeys optional and forces you to set a password already. If they switch their default to passkeys and add a password as an optional step, nothing would change.
My passkey is synched with iCloud which has all hardened settings set to on, so if I switch devices (I have) its seamlessly. You can also export Passkeys is my understanding but I have not tried it.
When I'm on non-Apple I can scan a QR code on my phone and it will validate it the same way, using bluetooth to send enough info to the target computer to get me logged on.
Bluetooth is actually only used to establish proximity. Both your phone and the browser connect to an internet relay, which lets the phone essentially pretend to be a USB security key connected to the computer.
My passkeys are in my self-hosted Vaultwarden. If I can't put them there, I don't use them.
It's not all sunshine and roses, though. Despite having Bitwarden set as the only passkey provider in my Android setup, the phone persistently only offers me Google. Which is empty, because as I said, I won't use one tied to things I can't control. Works great on desktops, though.
Passkeys can theoretically require you to be on hardware, I haven't found anything yet that requires that.
I think you can do passkeys wile having private key. When os has the private key its just more convenient way of doing passkey.
I think you can but it has to be supported by the website that you are using
I don´t think I've ever come across a service that only used passkeys. Username/email + password + 2FA is usually the primary form of verification. There's usually a way to recover your account through email.
I use passkeys only, for the backend admin app for my app.
They work fairly well. I use a login ID/PW to set up the passkey, then the passkey, for everything else.
Passkeys work well, but I feel as if they are still a bit too “fiddly” for your average mensch.
I’m pretty sure the private key is transferable, but Apple keeps them in the Keychain, so they seem locked into the OS. On my machine, I can also keep passkeys in 1Password.
A while back, I wrote up a series on implementing passkeys in iOS/MacOS: https://littlegreenviper.com/series/passkeys/
Doesn't that defeat the purpose? I thought the objective was to get rid of passwords
Brian Krebs' article is, in my opinion, a much better read for this story[0].
[0] https://krebsonsecurity.com/2026/09/fbi-probes-service-selli...
(2 days ago, 276 comments) https://news.ycombinator.com/item?id=49529621
Discussion https://news.ycombinator.com/item?id=49529621
We have too many non-technical people in charge of things who just make decisions based on politics and magical thinking about what is possible.
‘Just make the encryption secure and so we can read it’
‘Just check everyone’s id but make it totally secure’
They do not care about 'secure' part at all.
This is the answer. The more failures, the more justification for more draconian restrictions of civil liberties.
I can hear the defense now: "Oh, yeah, you blame the honest, good, handsome people trying their best to protect you and you let the hackers off scot-free! We must make sure that hackers don't have access to the tools that aid them to commit these crimes, like books and computers. Anyone could be a hacker."
That is an unfair conclusion. These people run complex networks like the rest of us, they probably have a range of detection systems and, also like the rest of us, an almost impossibly large attack surface to consider internally and on their supply chain.
The problem is that it is really, really hard to make something secure even if you try and follow all the best-practices you know.
I guess the awkward bit is marketing everything as certificate this, accreditation that and overselling how secure it is although I don't really know how else you would word it, "as secure as we know how"?
Those aren’t the people in charge. People like 93 year old Senator Chuck Grassley are calling the shots.
Funny was just testing the pilot of the Irish Government Digital Wallet. Definitely seems like the way forward if we're intent on doing identity verification. I'd rather the government mediate this than a bunch of random 3rd parties.
We also have a Danish wallet now, AltID, which implements an anonymized (assuming no collusion between issuer and eavesdropper or service provider) age verification protocol based on batches of single-use tokens which contain no personal information (except that they can be traced back to you by the issuer).
It's been released and in production since summer. Since then, several social networks have apparently started A/B testing age verification for their EU users, but how many of them actually integrate with the anonymous solution that is now available and in production? To my knowledge: 0. They all use Persona.
This highlights one of my main criticisms of EU's naive approach to regulation of tech companies. They fail to realize that any regulation that they impose will be complied with in the most malicious way possible, which is how we got cookie banners with dark patterns instead of a simple HTTP header saying no thanks to cookies.
In 2027 it will become mandatory for big tech to accept the EU digital wallets, so soonish they'll have to integrate with them.
Well, that's at least a positive. I hope this means they are also forced to accept the weakest credential type, and that they can't require you to present the identity proof that the wallet also provides when verifying your age. Because if they are allowed to do that, then that is what will happen.
But usually gov't will outsource to random 3rd parties, no?
In Finland they outsource the system to banks and telephone operators. Its a very strange system. As far as I know, its not possible to access government services just by being a citizen. You also have to have an account with one of these third parties to get in.
It's also impossible to use if you're an expat, and you get locked out of every Government online system that requires strong identification.
You can, but there's a couple extra hoops to jump through. You first have to go to the police station and get a state ID because banks don't accept US passports or Finnish residence permits for whatever reason. Then you can open an account at Nordea. Then you can go through Nordea's system for e-identification.
Expat...as in Finnish citizen living abroad.
One third party, managed by a public contract, seems much better than a parade of third parties for every service you interact with though right?
probably gov will outsource to 3rd party for gov to build system to track and manage ID. Sometimes though also to manage, as in Denmark's MitID mainly managed by NETS under government set rules.
As a citizen under the France Passoire[1] and in an increasingly fascist chauvinist nationalist drifting in the geopolitical landscape, I wouldn’t be that found of delegating too much of these responsibilities to some centralized governmental institutions.
Note that’s this is not here some rant against any governmental power, just that in context, large private group use them as puppets and shrink their budget which of course impact quality of deliveries (not shaming the people who do the hard job without the relevant means). And while more distributed governmental topologies would have their own caveats, at least it would less likely offer opportunities for single point of failure.
[1] https://francepassoire.com/
On the end of the day, it is the state that issues these ID documents. So if you let the government go bad, IMHO the form of the documents does not matter that much.
During the totalitarian communist rule in Czechoslovakia, the state would regularly interfere with passports of people considered not loyal enough - withholding them outright or inventing extra paperwork that was necessary for the border police to let you out of the country. They also controlled all supply of foreign currency, both in an out.
Then if someone was actually allowed to travel outside the country but failed to return, their family and relatives would be punished, including demotion at work & prohibition of higher education.
So if your government goes bad, this is what will happen - the form of the ID takes at that point does not make much difference.
I live in the UK and was having this exact discussion with someone recently - I'd actually prefer Apple to be the owners of my digital identity over the UK government who would happily throw you in jail for expressing support for Palestine Action.
I think this is a bit of a non sequitur. The government will still be able to throw you in prison in either scenario. They will also be able to compel disclosure of your records in either scenario. So the difference is really just that in the scenario where apple verifies your identity there's an additional actor in the mix over who you have zero influence. It seems strictly worse.
Your digital identity would then be under the jurisdiction of the US government, who (despite talking a big game) are arguably worse for that same sort of behaviour.
If anyone has to have this type of control, better it be a local national government that you can in at least some small way influence.
I understand the sentiment, but your government you can actually fix by voting.
If Apple (or another large international company) suffers from decreasing margins, gets a new CEO and decides to turn the data it sits on into money there is absolutely nothing you can do, and you might in fact still stay a "forced" customer because of network effects (=> just consider whatsapp being an important communication channel in many places worldwide).
I think this attitude in general is often harmful; if your government sucks, fix the government instead of making yourself dependent on some quasi-monopolist private company.
You can fix companies by voting a government too. In fact, governments are quicker at fixing companies than themselves.
> You can fix companies by voting a government too
Indirectly only. This is typically also always too late; instead of doing "the right thing" in the first place, companies are disincentivized by regulation from doing "bad things" again.
Regulations are like scar tissue, they don't help against getting burnt in the first place.
Preemptive regulation typically sucks, and is admittedly extremely difficult to get right; most governments don't even bother trying.
Corporation know this and exploit it ruthlessly-- there are almost never consequences as long as they keep to the letter of the law, even when acting with intent, against better knowledge and causing astronomical damage to society (just consider the whole leaded gas disaster for an extremely clear example).
Sure but even when it works it slower and less total than good policy. Of course, good policy is a rare bird, too.
> throw you in jail for expressing support for Palestine Action.
For those unfamiliar, you are, of course, allowed to peacefully protest in support of Palestine in the UK.
Palestine Action is a specific group that was controversially labelled as a terrorist group after they broke onto a runway and spray painted military planes.
https://en.wikipedia.org/wiki/Palestine_Action
It would be best for any such ID system to capture only the minimum data to fulfil their function. Pre-WW2 Dutch records also listed religion, which made it very easy for the Nazis to round up and murder them. Some examples from Kansas show that even recording sex/gender is risky - state law requires driving licenses to record the original sex/gender from the first birth certificate leading to trans women having an M sex/gender marker getting arrested and prosecuted for identity fraud.
LOL, no, not in this side thanks. Be it by giant transnational corp or centralized all powerful government, the ruin of the noosphere is equally awful.
> increasingly fascist chauvinist nationalist drifting in the geopolitical landscape
What's going on in France?
France has always been chauvinist and nationalist.
The principle reason why the Netherlands joined the EU was to sabotage the French-German alliance- all the alarm bells went off in the 1950s.
I wonder if I can buy my own driver license since I lost it and now I need a copy to get some paperwork done! Hackers please!
You know, it always bugged me that the NSA (and more directly Google, and my phone company, and so on) know where I was at exactly this time a year ago, but that I do not.
In Google Maps Timeline you can definitely see it (if you set it up and you brought your phone)
This is news to me. Never knew you could do that... too late now.
Isn't it turn on by default?
correct, timeline is opt-in.
I thought that's what LLMs are for?
If you are in California the DMV makes tens of millions of dollars a year selling all the data you give to the DMV, which is why I give them a P.O. Box.
CADMV claims on their web site that they cannot accept a P.O. box as a residence address. I have yet to find anything in California state law supporting this policy, though IANAL. Their enforcement seems to be quite lax.
It is a REAL ID requirement. It is federal law. States issuing REAL ID compliant identity documents must mail them to your physical address. USPS provides that data for address validation.
DPVCMRA = delivery point is a commercial mail receiving agent. Any sort of location with PO Boxes.
https://developers.usps.com/addressesv3#tag/Resources/operat...
Disclaimer: I used to work for my state's DMV.
As someone who spent years homeless off and on, I resent REAL ID discriminating against the unhoused, so I choose to use my privilege to reject it. I do not have or need REAL ID and my passport does not need a current residential address.
You cannot literally use "P.O. box" but if you use the virtual street address service the USPS offers now it works just fine.
Am I missing something? What do you mean the DMV makes tens of millions of dollars a year selling data to itself?
I used to work for my state's DMV. They'd sell vehicle registration data to various companies, which is how and why you get those "we've been trying to reach you about your car warranty" phone calls. I don't know about CA, but KY had a problem with tracking who ordered and who paid for that data. When I worked there, we found a number of "purchasers" who only paid for Year 1 but stopped paying afterwards.
Federal law requires state DMVs to supply that data to the car manufacturers. So if you own a Chevy, they have to send your data to Chevy in case there is a recall.
I suppose one other industry interested in this information is the bounty hunter industry (bail enforcement). I believe there are several platforms for licensed bounty hunters where they can receive this and similar information about a person/vehicle.
selling (data you give to the DMV) [to third parties], not selling (data you give) [to the DMV]
The DMV sells the data you give to the DMV. The DMV does not sell the data you give to the DMV to the DMV.
Haha, damn not enough caffeine this morning to parse correctly
In guessing you can't choose to opt out ?
your positional encoding vector seems a bit off :D
And again, there will be no monetary consequences for the companies that failed to secure our private data.
And governments will continue to force citizens to use these shitty companies for whenever they need id verification.
That's why I say "Our lobbyists have more money than your lobbyists". Every state has sunshine laws to show who the lobbyists are, what they lobbied on, and to whom. Some states separate those lobbyists into legislative & executive branch lobbying.
I suggest you look at who voted for those bills, who lobbied them and who hired those lobbyists.
And create new requirements normalizing id verification for increasingly mundane things assuring citizens are exposed to ever more breaches.
More like class action lawsuit, $500m settlement, $300m for lawyers and $0.50 for every victim.
The HN submission title is a garden-path sentence:
Hackers Had a Live Feed of Every ID Verification Company Scanned
(Huh? How do you scan a company?)
The original title is easier to parse:
Hackers Had A Live Feed Of Every ID This Verification Company Scanned
Thank you, it was very confusing indeed, the HN post should be fixed to something directly clearer
How exactly does that work? How can you sneak a live feed past detection systems? It is incomprehensible to me, considering this is highly regulated and sensitive data. It is just open ports sending what they shouldn't be sending all the way out or what?
Brian Krebs' article makes a good case for the ID source being a harvester on the internal Hertz Car Rental network, and likely other similar consumer services that log ID for asset security and recovery.
These are hardly military grade networks, as long as the driver licence scans make it to the database and can be used to identify and recover damages from accident or theft it's unlikely anybody has cared much past that functionality.
I’ve been following the development of the drivers license sharing system from Apple where different fields can be selected; are there any implementations of PKI based identification systems where multiple certificates can be generated and revoked when compromised?
I’ve often thought that replacing the US social security number with a more robust root key makes for a fun thought experiment. Hard to imagine how such a system could securely serve so many people but passports with embedded chips seem to be doing okay.
Not in the US. Several EU countries have PKI systems integrated with identity documents that let the requester to ask for age (for example) and then only age is supplied. But their PKI systems are for the whole ID document.
As for the passport, the key/PIN you need to authenticate to the chip are printed on the page with the photo. Otherwise "hackers" can only determine nationality of passport. The standard is ICAO 9303.
https://www.icao.int/publications/doc-series/doc-9303
SSN was never intended for identification. My original card, issued in the 1970s was clearly marked "not for identification". In the original numbering system, the first 3 digits identified the office/area where the card/number was issued and the next 2 digits identified the filing cabinet. 700s were set aside for railroad workers (until 1963) because the legislators did not want railroad workers to be included in social security.
https://secure.ssa.gov/poms.nsf/lnx/0110225045
No worries! Governments who used this company are taking responsibility and now have a plan to, at the very least, replace all IDs they forced people to expose and to make sure the old ones are unusable!
That's a sarcastic joke. It's how governments demand private companies react, but ...
> This week a massive new data breach has been revealed that should put the nail in the coffin for the idea that any sort of age or identity verification could be safe.
Yeah just like how the multiple breaches and utter negligence from the incumbent credit bureaus killed the credit file managed by private companies.
Crazy hack considering the order of magnitude...
"Nobody could have predicted this"
It's getting really tiresome
This is a sacrifice we just have to be willing to make as a society if we want to project kids from the horror of using the internet
The politicians really want to know who to prosecute if someone on the internet says bad things (about them).
I think there's a number of people reading this who clearly didn't detect the satirical nature of this single sentence. It's blunt and obvious, but even so...
I still don't understand why the simplest approach isn't used: ban kids from using the Internet unsupervised. There's really no good reason why a six year old should have internet access.
The argument is that there are parents who are too stupid/lazy to enable parental controls on kids devices and society has a duty to protect kids even if their parents are negligent. Also, kids interact with other kids, so even if you do everything right your kids wind up with access/peer pressure through the kids with bad parents.
I dunno if I agree but I think that's the thrust of it.
it's pretty easy to enforce that for 6 year olds. not as easy for 12 year olds.
This is precisely why the authority doing these checks needs to be the government that already issues the IDs.
Using ZKP as the EU proposes is the only way to prevent this data being leaked to unreliable third parties and leaves the knowledge with the institution it derives from in the first place.
I don't know why HN rails against it constantly, it is the obvious technical and organizational solution to this issue.
[delayed]
Government systems leak information all the time. The type of institution managing the data makes little difference. Its how the institution manages the data that matters.
But the government inherently has that data, as it comes from there. They're the ones issuing the IDs in the first place.
Theres no avoiding this, structurally. So the best thing you can do is not to introduce any additional points of failure.
All previous systems avoided this. The government issued me an ID in the past, yet had no record of when I used it, or for what.
I am so much more afraid of monopolies invading my privacy than roving hackers, or my corner store. Governments are the ultimate monopoly.
The government doesn't know what you use your ID for in a properly implemented ZKP system, either.
And clearly, the alternative is not the corner store. It's private monopolies building databases of 153M IDs.
It is European which means that it is both anti-capitalist and communist and therefore must be spit upon.
That was sarcasm.
Are you sure? It’s really hard to differentiate nowadays
> Are you sure? It’s really hard to differentiate nowadays
Case in point; I can't tell if you're being sarcastic or not! :D
https://en.wikipedia.org/wiki/Poe%27s_law
I cant event tell if you are being sarcastic or not :)
All the kids will be safe now they're logging into porn sites as Pete Hegseth.
Only after they've had their mandatory scrotum inspection and testosterone check to join the military at age 18.
You guy don't have to show cock and balls to the military commission? It's a standard practice in post-Communist countries, including NATO ones.
In the olden days, the doc would hold your nuts and tell you to turn your head and cough. This was to detect if you had a hernia (which was a medical exception).
The bunghole inspection was to see if you were the receptive partner in anal intercourse because being a gay man was a criminal offense.
Allegedly, lesbianism was never outlawed because Queen Victoria was so offended by the idea of it that she prohibited the legislation from even being heard in Parliament.
https://en.wikipedia.org/wiki/Royal_assent
How do you even conclude from inspecting a butthole that it is "a receptive one"? Every man was one butthole inspection away from being a criminal?
I mean if a poo can get out, something of its size can also go in, and quality toilet paper was not that available at all historically.
There's a guy in the US military intake processing facility (MEPS) whose job is, among other things, to inspect your bunghole
https://www.google.com/search?client=firefox-b-d&q=MEPS+assh...
But more serious and non joking answer, the new thing from the "department of war" is testosterone level lab exams for existing servicemembers.
https://news.google.com/search?q=US%20military%20testosteron...
so... for military grade identification systems instead of face photo with id document next to it, you make banghole photo with id document next to it?
It's all good and fine until a scan of your combined face ID / bunghole ID is for sale on a russian based internet crime forum.
https://www.google.com/search?num=10&client=firefox-b-d&hs=Y...
Except this helps no child.
Isn't that the one that was left behind in 2002?
> There is no safe age verification. There is no age verification that doesn’t put people at risk.
There are zero knowledge proofs
True. I built a ZK age verification based on Polish digital identity https://x.com/maciejlotkowski/status/1899896737688436844, but I didn't find a business case for it at the time.
There's a EU initiative https://digital-strategy.ec.europa.eu/en/news/commission-mak.... The direction is generally good, but I'm not very positive about the implementation (as with everything comes from the govs).
More information here about how the implementation enforces a Google/Apple duopoly: https://github.com/eu-digital-identity-wallet/av-doc-technic...
Concrete ZKP age verification schemes are hardly zero knowledge.
Imagine your idealized ZK address verification scheme. It would go something like: I show up at a website, it sends me some challenge, I send back a signature of the challenge that could only be made by someone with an of-age ID, but without specifying who. Everyone is happy.
Now little Johnny borrows my ID, and uses it to setup some oracle that provides ID validation for every kid and bot in the country. Woops.
To stop that you must compromise the idealized zero knowledge properties of the scheme, and in doing so you create the potential for harm/risk for everyone.
Sure, it's better than sending an ID card live feed to the dark web, but the risks of ID card theft are at least somewhat easy to understand.
Some of the threats to human rights don't even require the departure from the 'idealized' model-- as even the idealized model requires an ID issuer to issue the of-age person an ID. And so if the ID ZKP is widely required then the issuer can unperson you by simply declining to issue you an ID.
add MFA to the check
To where, the site requesting the verification? Now it is no longer zero knowledge.
No, to the ID to prevent abuse if the card get stolen.
add a different ID check to the MFA if that doesn't work, then add more MFA to that new ID check. Eventually it has to work, right?
It's definitely worth doing infinite security in order to avoid regulating social network algorithms, because
To prevent abuse add MFA to the ID. Problem if stolen cards solved and still zero knowledge.