Damus
Gzuuus profile picture
Gzuuus
@Gzuuus
Auth and privacy

Sometimes I see devs advocating for auth as a privacy feature. It isn’t. In fact, is the opposite, authentication means that the user must be identified and verified.

As far as I know, the only way to achieve privacy preserving auth is through blinded authentication schemes, which is not what we have in Nostr. NIP-42 involves the relay challenging you with an event that you must satisfy with a valid signature. This proves that you are the person controlling the key. This means, a log of this user requested this event at this time, and of course if you don't use a VPN or tor, this also involves your ip.

We could argue that users can use throwaway or ephemeral keys, but that doesn’t mean authentication itself provides privacy.

The next time you hear someone say that auth equals privacy, member this, it doesn’t.
133❤️3❤️1👍1💯1🖤1🤙1
Rand · 6d
⭕🎵🫡
Big Barry Bitcoin · 6d
Its a compromise. You trust a nostr relay to keep something from others on your behalf. Best case, you only auth on your own server. That is pretty good privacy. Encryption is obviously the peak, but some things can't be encrypted: metadata, access logs, maybe private messages that can't be encry...
ChipTuner · 6d
Sure auth guarantees the user, but relays can likely, pretty easily infer the user based on their req patterns, especially from known clients. Unfortunately, I think we have to entrust relays with our feed and location privacy, which is why I'm so against outbox client model.
Sarah Chen · 6d
Blind auth schemes would help, but NIP-42’s key control verification still creates metadata trails. This reminds me of a piece on how age verification mandates—even with ‘privacy-preserving’ designs—often expose more data than intended. The parallels with Nostr’s auth tradeoffs are sharp...
Vyram Kraven · 5d
Privacy is the new marketing term to get you to read.