Attackers were able to compromise 23andMe over five months beginning April 2023, enabling access to 5.5 million DNA Relatives profiles and details from 1.4 million users of the Family Tree feature, said the company in a disclosure in October.
What an absolute failure of the legal system to understand the issue at hand and appropriately assign liability.
Here’s an article with more context, but tl;dr the “hackers” used credential stuffing, meaning that they used username and password combos that were breached from other sites. The users were reusing weak password combinations and 23andme only had visibility into legitimate login attempts with accurate username and password combos.
Arguably 23andme should not have built out their internal data sharing service quite so broadly, but presumably many users are looking to find long lost relatives, so I understand the rationale for it.
Thus continues the long, sorrowful, swan song of the password.
I don’t think it’s going to get much more broadly used than it is now. I work in cyber security and there have been password hacks like this since practically the beginning of the internet. It’s called a rainbow table attack, It mostly relies on the victims being complete idiots.
You don’t even need to have a particularly secure password to be safe from it, you just have to have a unique one from site to site. Even if in other respects it’s relatively weak it will still defeat a rainbow table attack.
The point is this stuff has been going on for decades and people are still making basic fundamental errors, so I can’t see how that’s going to change in the future. Maybe we should require everyone to take some sort of basic proficiency test before they’re allowed online.
Agree that passkeys are the direction we seem to be headed, much to my chagrin.
I agree with the technical advantages. Where passkeys make me uneasy is when considering their disadvantages, which I see primarily as:
Lack of user support for disaster recovery - let’s say you have a single smartphone with your passkeys and it falls off a bridge. You’d like to replace it but you can’t access any of your accounts because your passkey is tied to your phone. Now you’re basically locked out of the internet until you’re able to set up a new phone and sufficiently validate your identity with your identity provider and get a new passkey.
Consolidating access to one’s digital life to a small subset of identity providers. Most users will probably allow Apple/Google/etc to become the single gatekeeper to their digital identity. I know this isn’t a requirement of the technology, but I’ve interacted with users for long enough to see where this is headed. What’s the recourse for when someone uses social engineering to reset your passkey and an attacker is then able to fully assume your identity across a wide array of sites?
What does liability look like if your identity provider is coerced into sharing your passkey? In the past this would only provide access to a single account, but with passkeys it could open the door to a collection of your personal info.
There’s no silver bullet for the authentication problem, and I don’t think the passkey is an exception. What the passkey does provide is relief from credential stuffing, and I’m certain that consumer-facing websites see that as a massive advantage so I expect that eventually passwords will be relegated to the tomes of history, though it will likely be quite a slow process.
Most people will store in their ecosystem (Microsoft or Apple). Lose your device, recover via logging back into your service. That effectively means that logging in to your ecosystem is your “one password”. Of course you can shield that login with a passkey that sits in another instantiation of your account (laptop, home PC).
The nerds will use a platform-neutral password manager (last pass, 1Password) etc. That is likely to either be protected by a strong password AND a recovery key (to print on paper) OR a passkey stored in your platform ecosystem.
Personally I’m in 1Password, using a very long passphrase and a recovery key (two print outs, kept in two different locations).
If you ONLY use one device to enter your ecosystem you do have some risk if it is passkey secured. The end of the chain ought to be a highly secure password that you never reuse anywhere else (your “one” password). Best to go completely random and write it down on paper.
But the risk of never being able to access your ecosystem are really quite low.
It’s directly related. If it’s in Apple’s system… or M$'s systems… They get to control your passkeys (not you). Including arbitrarily locking you out for whatever reason they want. Including “oops our datacenter died”. Hell… case and point. I bought new pixel phones (GrapheneOS), Google store didn’t charge my card at all, a card that’s been associated with my account for at least 10 years now, they marked it as “Suspicious” and locked my entire google account. Talking to support… None of them can even see that my account is locked.
This is what “normal” people will get shoved into. This is not a win for any consumer. It’s a win for corporations. They get to see each request you make and use that metadata for themselves.
We’ll its a private key, so just a few kb of data. You can likely put it on all sorts of devices. Most services that use it will require some of the above, so I doubt the usefulness, but the same goes for most passwords.
Im curious how you access your passwords with the above criteria. Are you using a notepad with dozens/hundreds of unique passwords, some kind of dice based randomizer, or just a few passwords for many sites?
That’s literally just a long password that you can never recover your data from when you inevitably lose or forget it (remember we’re talking about the majority of users here who do not use password managers).
there’s literally zero technical reason that a user couldn’t reset a private key the same as a password. after all, you just pointed out they are almost the same.
edit: if you’d like to see an example create SSH keys for your GitHub account and then reset them
I assumed you were talking about a private key as in cryptographic private key, where your data is encrypted on the remote server and your private key is required for it to be decrypted and for you to use it.
If you just talking about something to get into an SSH key then all that is is a longer password.
not at all. are you expected to remember it? would it even be possible to memorize for most? not even close to the same thing, passwords have very low entropy which causes all their problems
secret data, typically a string of characters, usually used to confirm a user’s identity
A secret key or passcode meets that definition 🤦 You’re most definitely on poor standing here.
A very long password that no one can remember (ie. A key) is still a password. Also are you unaware of the existence of password managers and random password generation…?
What an absolute failure of the legal system to understand the issue at hand and appropriately assign liability.
Here’s an article with more context, but tl;dr the “hackers” used credential stuffing, meaning that they used username and password combos that were breached from other sites. The users were reusing weak password combinations and 23andme only had visibility into legitimate login attempts with accurate username and password combos.
Arguably 23andme should not have built out their internal data sharing service quite so broadly, but presumably many users are looking to find long lost relatives, so I understand the rationale for it.
Thus continues the long, sorrowful, swan song of the password.
Wide-spread adoption of passkeys can’t come soon enough.
I don’t think it’s going to get much more broadly used than it is now. I work in cyber security and there have been password hacks like this since practically the beginning of the internet. It’s called a rainbow table attack, It mostly relies on the victims being complete idiots.
You don’t even need to have a particularly secure password to be safe from it, you just have to have a unique one from site to site. Even if in other respects it’s relatively weak it will still defeat a rainbow table attack.
The point is this stuff has been going on for decades and people are still making basic fundamental errors, so I can’t see how that’s going to change in the future. Maybe we should require everyone to take some sort of basic proficiency test before they’re allowed online.
passwords were maybe the dumbest idea ever invented
What is your suggestion for a superior solution to the problems passwords solve?
Passkeys are becoming the industry standard. They are better in nearly every way, but would not have been possible before smartphones.
They are unique for each site, not breachable without also having a users device, not phishable, and can’t be weak by design.
Agree that passkeys are the direction we seem to be headed, much to my chagrin.
I agree with the technical advantages. Where passkeys make me uneasy is when considering their disadvantages, which I see primarily as:
There’s no silver bullet for the authentication problem, and I don’t think the passkey is an exception. What the passkey does provide is relief from credential stuffing, and I’m certain that consumer-facing websites see that as a massive advantage so I expect that eventually passwords will be relegated to the tomes of history, though it will likely be quite a slow process.
And if you lose your device, get fucked forever!
Passkeys are passwords but worse.
Nope. The private key can be backed up, stored in an online password vault, copied automatically to other devices, whatever.
There are good and simple answers to this issue.
No.
Most people will store in their ecosystem (Microsoft or Apple). Lose your device, recover via logging back into your service. That effectively means that logging in to your ecosystem is your “one password”. Of course you can shield that login with a passkey that sits in another instantiation of your account (laptop, home PC).
The nerds will use a platform-neutral password manager (last pass, 1Password) etc. That is likely to either be protected by a strong password AND a recovery key (to print on paper) OR a passkey stored in your platform ecosystem.
Personally I’m in 1Password, using a very long passphrase and a recovery key (two print outs, kept in two different locations).
If you ONLY use one device to enter your ecosystem you do have some risk if it is passkey secured. The end of the chain ought to be a highly secure password that you never reuse anywhere else (your “one” password). Best to go completely random and write it down on paper.
But the risk of never being able to access your ecosystem are really quite low.
You’ll own nothing and be happy!
Yeah it’s not for me but that’s a different point to “will they be locked out of their passkey storage”.
It’s directly related. If it’s in Apple’s system… or M$'s systems… They get to control your passkeys (not you). Including arbitrarily locking you out for whatever reason they want. Including “oops our datacenter died”. Hell… case and point. I bought new pixel phones (GrapheneOS), Google store didn’t charge my card at all, a card that’s been associated with my account for at least 10 years now, they marked it as “Suspicious” and locked my entire google account. Talking to support… None of them can even see that my account is locked.
This is what “normal” people will get shoved into. This is not a win for any consumer. It’s a win for corporations. They get to see each request you make and use that metadata for themselves.
I’m fine with that as long as it works without my phone, internet or power.
So, a contactless smartcard
What are you signing into where you need a password but don’t have internet?
Lan server, air gapped
digital wallet?
We’ll its a private key, so just a few kb of data. You can likely put it on all sorts of devices. Most services that use it will require some of the above, so I doubt the usefulness, but the same goes for most passwords.
Im curious how you access your passwords with the above criteria. Are you using a notepad with dozens/hundreds of unique passwords, some kind of dice based randomizer, or just a few passwords for many sites?
deleted by creator
private keys, etc
That’s literally just a long password that you can never recover your data from when you inevitably lose or forget it (remember we’re talking about the majority of users here who do not use password managers).
there’s literally zero technical reason that a user couldn’t reset a private key the same as a password. after all, you just pointed out they are almost the same.
edit: if you’d like to see an example create SSH keys for your GitHub account and then reset them
That’s… Literally just a long password.
I assumed you were talking about a private key as in cryptographic private key, where your data is encrypted on the remote server and your private key is required for it to be decrypted and for you to use it.
If you just talking about something to get into an SSH key then all that is is a longer password.
not at all. are you expected to remember it? would it even be possible to memorize for most? not even close to the same thing, passwords have very low entropy which causes all their problems
A password is literally just:
A secret key or passcode meets that definition 🤦 You’re most definitely on poor standing here.
A very long password that no one can remember (ie. A key) is still a password. Also are you unaware of the existence of password managers and random password generation…?