Comment on Sunlight creates quantum entanglement once thought to require lasers
natanael@slrpnk.net@slrpnk.net 2 weeks ago
This is neat, but quantum key distribution is stupid as hell because it’s only secure if you START OFF WITH A SHARED SECRET to authenticate the key pad and first round of messages.
And if you have that then you don’t need quantum encryption because you can do everything with just symmetric encryption (unless you specifically have some very weird forward secrecy requirements where it isn’t approved to use asymmetric post quantum algorithms like McEliece)
threelonmusketeers@sh.itjust.works 1 week ago
Wait, really? There’s no way to establish a secure connection over an insecure channel via quantum encryption?
natanael@slrpnk.net@slrpnk.net 1 week ago
Yeah that’s the thing, quantum key distribution can not prove who is on the other end!
It fundamentally can not guarantee that the line is intact before you start communicating. You can in theory detect a break that starts after, but one that happened before is effectively invisible.
Quantum key distribution can not do anything other than key expansion (giving you more secret bits from s few), because the only way to know who is in the other end is to have agreed to a method of validation, such as a signature algorithm (but if you trust signatures you won’t care about quantum key distribution) or a shared secret (an authentication tag key). But if you have a large enough secret key, you can just use a regular symmetric encryption algorithm. Some might want forward secrecy from the quantum line, but you can get that with symmetric key ratchets like Signal uses where the ONLY benefit of quantum key distribution is that you avoid key management (but instead you maintain physical infrastructure)