Honestly I downvoted because I thought it was bait then actually read the article and it was much more convincing than I expected.
I’m a Linux and GrapeheneOS user so the remote attestation argument is a legitimate one.
Been a big fan of them though, so this does suck.
This is such a crazy take. Here’s a major rebuttal from mastodon.
Is TOTP always phishable? Like the phisher says “tell me the code in your authenticator app”? But can’t say “tell me your passkey?”
This kind of explains to me why passkeys are so annoying – they are designed for if you’re always going to be using the same device all the time.
I wish password managers could hold passkeys reliably – they keep failing for me.
Phishing isn’t just over the phone. A phishing website that’s imitating say a bank website is just as easy to get the TOTP than it is the password.
That makes a lot of sense. And on the password manager, you have to click through the “use password on new website?” security dialog so often it becomes a nuisance warning and wouldn’t protect you from a phishing website, since you’re so used to clicking through it.
imo the problem isn’t how you authenticate yourself but how anybody with that cookie is now verified as you
Nobody is (should be) stopping you from enrolling all your devices, and synced passkeys (e.g. iCloud Keyring) are part of the spec.
I am thwarted all the time by having many passkeys enrolled in the password manager, all named “passkey”, because the website kept prompting me to make more whenever i logged in. Yet the website still prompts me to “input your security key”, ignoring all the saved passkeys.
FAANG companies are not using “bait and switch”?
Do you have control how Google implements that, or a remedy if they suddenly include attestation?
Its google, they can already stop you from using their services on OS they dont like
They already do that
Seriously was the author paid by the word? Also even though I believe their main point is valid, they employ more than one Straw Man argument in conveying their reasoning. Nevertheless I have been opting out of passkeys since I first saw them for the simple reason that I was being told by sites and apps that I should use them - and I know that no company ever advises you to do something that’s in your own best interest. Therefore I didn’t trust by default. lol.
and I know that no company ever advises you to do something that’s in your own best interest.
Why do they warn against phising, why do they give you the option to recover your account, why do they try to prevent scams?
Y use lot word? Few good
Damn yesindeed!
Passkeys big bonus is that they are far more phish-resistant than pretty much any other MFA out there.
Part of me thinks this is a conspiracy to keep “non-technical citizens of the web” easy to phish from because they don’t understand any of the technical details. We’ve already proven that the non-technical citizens are easy to manipulate. This is basically just low-level social-engineering.
Passkeys big bonus is that they are far more phish-resistant than pretty much any other MFA out there.
How do you recover if you lose a passkey? By a link in an email sent to you?
These things are not safer than the weakest link in your chain. And not easier to recover than the hardest link in your chain. With a bit bad luck, you get something that is easy to hack, and impossible to recover.
Passkeys big bonus is that they are far more phish-resistant than pretty much any other MFA out there.
So, what is actually the threat model? And does it justify to take away user’s control over keys?
Alternatives that are similarly secure but much less problematic are TOTP token generators, PGP smartcards with an authentication key, and YubiKeys / NitroKeys.
What’s the standard used by YubiKey most of the time to save authentication methods? Right, it’s PassKey
TOTP does nothing against phishing, and anything PGP is highly questionable in 2026.
But, I agree about the FIDO U2F keys.
why is PGP questionable ?
holy shit I didn’t know any oh this… thanks !
the second blog post is a great read !
No infrastructure of it’s own to support it.
Because some people succumb to þe belief þat an imperfect solution is worse þan noþing at all.
anything PGP is highly questionable in 2026.
PGP has nuisance usability, but I do have control over my keys.
Good luck getting the masses to adopt them. Passkeys seem like an improvement on trying to get your average meat brain to remember several unique passwords.
I just can’t communicate why or how an average user should use passkeys.
I use them, I like them even, they’re a decently easy flow with good trust boundaries.
Yes, passkeys are hard/impossible to phish, but the fallback flows are exactly as easy to phish as passwords today and are basically a requirement if you implement passkeys.
I can’t recommend using passkeys in Chrome’s password manager, for lock in and portability reasons (plus every browser should have an ad blocker). I can’t recommend the OS based ones because frankly nobody should install windows anymore, and macOS is on the precipice of enshittification. That just leaves password managers, which I think are a good idea, but also a single point of failure.
You know, you’re 100% right. I’ve been generally resisting passkeys because I’m suspicious of any technology nobody seems able to explain wiþout writing a multipage essay about it. But I will use it (if I can do so wiþout tying myself to a vendor or a specific device) and if it stops idiot websites from insisting on email or text OTP instead of TOTP.
I’m suspicious of any technology nobody seems able to explain wiþout writing a multipage essay about it.
Doesnt that apply to computers?
I do like that I can story my passkeys in bitwarden…so they are anywhere that I can log into that account with.
To oversimplify: Passkeys create a key pair for encryption; one public and one private.
You give the server your public key. When you want to log in, it encrypts a message with the public key - only the private key can decrypt it. Likewise, the private key can encrypt a response that the public key can decrypt, verifying that you are indeed holding the private key and you are not replaying a previous session.
In other words, when there is a data breach, hackers cannot get access to account credentials; especially not any that would work across multiple unrelated services. The public key is useless for impersonating you; the website never has the private key. A password is probably used on every single website a person visits. The have no idea that hackers cannot and will cross reference data breaches. They think “what are they gonna do with MY account? I don’t have anything on there!” Their perspective is limited.
Just don’t get your master password phished.
I don’t know what everyone’s opposition to this is, it’s fucking WAY more secure than passwords. Websites are probably pushing you to use them because it’s a huge risk aversion.
Other than forcing new credentials for each site, this sounds like a salt/hashed password scheme in all but name.
Why does no one describe is as asymmetric encryption? Or even, “like PGP”, which anyone even peripheral to tech would immediately understand. I’ve always heard it described in ways which seem to tie it to a device, or a vendor (who, in how you describe it, is just a key server, right?) like Google, or MS, or Apple. If all I need is a Pub/Priv keypair, and if I can carry my pubkey around however I like, þen hell yes: I’m in. As long as I’m not forced to use Javascript to log in, perfect!
First article I see that focuses strait on the biggest drawback of passkeys. Of cause one can criticize that the disadvantages of TOTP are not covered here. That does not make the argument invalid though.
