News for hackers. Stuff that matters.
Re:YubiKey Bio Series by amelius on Tuesday October 05, 2021 @05:32PM (#28765900)
Can someone explain why this makes authentication more secure?

If I tap the device while a malicious process is attempting to access another machine in the background, what good does this device do?

Re:YubiKey Bio Series by jolmg on Tuesday October 05, 2021 @05:39PM (#28765982)
Not with the Bio series apparently, but at least on the Yubikey 4, when you're setup with OpenPGP, a tap works with a single request and it prompts you to tap with a blinking light. If it's blinking when you haven't done anything to prompt a tap, that would be something to worry about. If you do something that prompts a tap and a malicious process also requests something from the key at the same time, you'll get 2 requests. You'd tap it for the request you're expecting, see it blinking faster to confirm your tap, and then blink slowly again to request the 2nd tap. That second request would be your cue for worrying. IOW, the blinking is an alert against your scenario.
Re:YubiKey Bio Series by amelius on Tuesday October 05, 2021 @05:52PM (#28766141)
Ok, thanks! How do well does this key go together with workflows that use automation (e.g. daily rsyncs over the network that run in the background)?
Re:YubiKey Bio Series by jolmg on Tuesday October 05, 2021 @05:59PM (#28766222)
Not sure how you want to combine automated daily rsyncs with hardware keys. You can use OpenPGP for ssh authentication, and set it up so it prompts you when rsync starts automatically. However, do you really want your automated jobs to depend on having a human to answer prompts or at least having a key inserted when it's scheduled to run? It sounds like there could be a better setup.
Re:YubiKey Bio Series by amelius on Tuesday October 05, 2021 @07:07PM (#28766891)
My issue is that once I start using authorized_keys for automated processes, then it defeats the purpose of the key.
Re:YubiKey Bio Series by jolmg on Tuesday October 05, 2021 @07:14PM (#28766940)
Why? authorized_keys holds public keys. Why would it defeat the purpose of a hardware private key?
Re:YubiKey Bio Series by amelius on Tuesday October 05, 2021 @07:23PM (#28767022)
Because if I can automatically log in through authorized_keys (used by my scripts), then it doesn't make sense to also have an authentication method that requires a key.
Re:YubiKey Bio Series by willis936 on Tuesday October 05, 2021 @10:40PM (#28768559)
You could make an account with restricted permissions then use ssh-add for automated authentication.
Re:YubiKey Bio Series by staticassertion on Tuesday October 05, 2021 @06:46PM (#28766697)
FIDO keys are really more about avoiding human errors/ human auth. If you need to authorize a server you may want to just use a client certificate.
Re:YubiKey Bio Series by karmakaze on Tuesday October 05, 2021 @06:25PM (#28766506)
I don't even understand the threat model for this, and why it's better than the fingerprint sensor built into the laptop. Seems like it's just writing the software elsewhere. What's up with the micro Yubikeys always left in a usb port operable by anyone made even less sense, so a step up.
Re:YubiKey Bio Series by g_p on Tuesday October 05, 2021 @06:40PM (#28766642)
Compared with a button based yubikey, someone else can't trivially tap this one if it's left behind, as you say.

From a threat model perspective, biometrics are far from perfect. This setup avoids disclosing the biometric to the host system though - the yubikey itself validates the biometric and signs (or doesn't sign) a response with an internal protected private key. If you had a sensor built into a laptop or workstation, you are trusting that sensor. Some old ones used to expose raw fingerprint data to the operating system, which is clearly bad.

If the user brings their own reader, you don't need to expose any biometric information beyond the yubikey itself (which is issued per user), so a host compromise or a rogue system won't be able to compromise someone's biometrics. Admittedly in this case a lot of the threat models that make sense would involve shared computers etc, and those have their own security issues.

Re:YubiKey Bio Series by yoz-y on Tuesday October 05, 2021 @06:30PM (#28766554)
The threat model is that people can't ssh to your machine in order to elevate their access (e.g. corporate VPN), because they can't physically touch the key. Example: whatever happened when major Twitter accounts were compromised.
Re:YubiKey Bio Series by karmakaze on Tuesday October 05, 2021 @08:48PM (#28767744)
Yes, but how is that different than if the software integrated with the sensor built into the hardware, like when 1Password uses the laptop's fingerprint reader?
Re:YubiKey Bio Series by yoz-y on Wednesday October 06, 2021 @12:10PM (#28774804)
There are some technicalities (I'm not sure what they are, but some orgs may want to have some boxes ticked for corporate approval of a tool). Then there is the fact that you can actually use the key in multiple machines, even those that don't have biometrics (e.g. desktops)

Personal use case: I have my 1password account locked with a YubiKey, which means that if somebody would steal my credentials they can't get in without one of my devices. However, I still can technically use it from a browser on a friend's computer if I really needed to.

Re:YubiKey Bio Series by cwp on Tuesday October 05, 2021 @05:49PM (#28766111)
I think the point of this is that if somebody steals your YubiKey, they also have to steal your finger to be able to use it.
Re:YubiKey Bio Series by jolmg on Tuesday October 05, 2021 @05:51PM (#28766131)
If this had OpenPGP support (though it may also be true of other modes) they'd also need to know your PIN code.