Re:Quantum Resistance and the Signal Protocol
by simcop2387 on Tuesday September 19, 2023 @03:46PM
(#37575421)
> NSA interventions in cryptographic design have historically gone in the opposite direction[1].
I'm not sure I'd say that given that there are some other designs and things that have gone on[1][2]. Particularly the Dual EC debacle. They have a history of helping make suspect or down right compromised crypto if they think they can get away with it. That said it does look like they avoid doing it to anything that gets USA GOV approval for use internally but it's difficult to say to what level they would actually go to for getting a backdoor out into the world that would let them look at other secrets.
[1] https://en.wikipedia.org/wiki/Export_of_cryptography_from_th... [2] https://en.wikipedia.org/wiki/Dual_EC_DRBG
Re:Quantum Resistance and the Signal Protocol
by woodruffw on Tuesday September 19, 2023 @04:07PM
(#37575766)
That’s fair. Maybe this is too fine of a hair to split, but I would categorize the Dual_EC fracas as less an intervention and more of a ham-fisted attempt to standardize something that mainstream cryptography was immediately suspicious of. But I suppose you could argue that there was similar suspicion around DES from the very beginning.
The same will be true of PQ signing schemes, meaning that a backdoor would be predicated on the USG believing that they have some truly remarkable NOBUS breakthrough in PQC. That feels unlikely to me; NSA interventions in cryptographic design have historically gone in the opposite direction[1].
(This is separate from the actual design question. I don't know as much about stateless PQ signing schemes, but the stateful ones are mostly "boring" hash-based cryptography that's well understood to be strong in both settings.)
[1]: https://en.wikipedia.org/wiki/Data_Encryption_Standard#NSA's...