4 ms·
Tangential forward-looking paranoia: I've got to wonder if the DUAL_EC debacle only appears so ham-fisted because the public understanding of public key crypto
by mindslight 11y ago
Tangential forward-looking paranoia:
I've got to wonder if the DUAL_EC debacle only appears so ham-fisted because the public understanding of public key crypto is much further ahead than our understanding of symmetric ciphers. Universities employ armies of mathematicians studying mathematical structures for their own right, whereas shuffling bits isn't sexy.
Conversely, "Clyde Frog" has been studying symmetric ciphers much longer and harder (symmetric is sufficient for nation-state security) and could have a deep symbolic understanding of common symmetric constructions akin to how we see the public-key math. They would then know how to choose constants that admit similar backdoors, and the entropy of "nothing up my sleeve numbers" isn't exactly well quantified.
Rather than a proactive attempt, DUAL_EC could have been a reaction to worries about movement to RNGs based on asymmetric math.
- why-el 11y ago> symmetric is sufficient for nation-state security Is that a verifiable assertion? Would love to read more about it.
- mindslight 11y agoGovernments can afford to employ people to transport N^2 keying material. Additionally, they have predefined communications patterns. Even if part of the government moves to asymmetric algorithms for key distribution (only possible after the discovery of Diffie-Hellman), the top secret portions can continue using couriers to avoid relying on an additional algorithm. Combined with its standard use for bulk ciphering, symmetric is obviously the more valuable target to secure/break.
- fossuser 11y agoPublic Key crypto was discovered by GCHQ (and then given to the NSA) several years before it was publicly discovered by Diffie-Hellman and RSA. I think this was to avoid having to have the symmetric keys under armed guard. The public discovery is also what kicked off the 'crypto wars'. I'd be surprised if modern nation state intelligence communities found symmetric encryption sufficient. A lot of interesting information about the history I learned from Steven Levy's crypto: http://www.amazon.com/Crypto-Rebels-Government-Privacy-Digital/dp/0140244328 http://www.amazon.com/Crypto-Rebels-Government-Privacy-Digit...
- mindslight 11y agoOh yeah, I'd forgotten about Cocks. Blame my selective memory for withholding credit from people who don't share. Spooks would of course welcome any discovery, and asymmetric crypto does solve problems for them (getting government crypto distributed as wide as possible). I am saying purely symmetric is "sufficient" for their core functionality - the communications that really need to be secret. Coupled with the head start before asymmetric was even discovered, that is where their focus is going to be. Put another way: if you were in charge of securing communications and had to prioritize resources, would you rather research a trustworthy asymmetric algorithm or a trusty symmetric algorithm? Likewise if you wanted to snoop on others' communications, would you prioritize breaking symmetric or asymmetric techniques?