Post #3243862

2025-10-15 14:37 UTC

@jaybird110127@dragonscave.space @micr0@wetdry.world Replay attacks are one of the most basic cryptographic protocol weaknesses. Hence my suggestion to use PBKDF derived asymmetric signatures for signing the reply to the random verification token. By randomizing the verification token replay attack is mitigated, and by signing with PBKDF derived signature impersonation is mitigated.

Replies (1)

@jaybird110127@dragonscave.space 2025-10-15 14:43

@datenwolf@chaos.social @micr0@wetdry.world Still, if an adversary has access to the user's devices, they could potentially respond to the ping in exactly the same way as the user would have, thus failing to communicate that anything is in fact wrong. What would really be needed is some form of biometric authentication which cannot be replicated or bypassed, but of course not every user is going to have a fingerprint sensor, etc. on their devices. Even my idea of a passphrase is susceptible to coercion, unless there's also a panic passphrase designed to be sacrificed, which acts just like the real thing, but in fact sends a panic notice to emergency contacts. I've never actually been in such a situation, but I've always been fascinated by systems intended to deal with such situations.

Open ##3243863