3 ms·

If I were a Project Lead at NSA's backdoor division, I'd also make it look like when the backdoor is discovered (vulnerabilities are always discovered, even if

by mtgx 7y ago

If I were a Project Lead at NSA's backdoor division, I'd also make it look like when the backdoor is discovered (vulnerabilities are always discovered, even if it happens 2 decades later) it will look like a "honest mistake" or even a "dumb mistake" that a developer made. Definitely wouldn't call backdoor variables something like "rear_entry", "remote_auth", etc. I don't assume all NSA spies are smart enough to think like that and sometimes NSA code may be a dead-giveaway, but I do assume that at least some of them are, for the more important/bigger backdoor projects. You can also look at NSA's history with IPSec, or the more recent Simon & Speck situation, where they pushed forward their own design that at first glance might seem fine, but if you tried to push them on the details, they would stonewall you.

yodon 7y ago
The NSA also didn't want to explain their reasoning for shortening the DES key size to 56 bits. It took many years for the public crypto community to discover the technique called differential cryptography, at which point it became clear that the actual theoretical maximum strength of the DES crypto system was 56 bits and the originally proposed longer key sizes didn't actually offer any true strength over 56 bit key variants. It remains important to be distrustful of all things in cryptography, but it also remains important that the one time the NSA had a huge and apparently negative impact on the strength of a crypto system, they were actually not weakening the system at all they were instead providing the public community with a significantly more accurate measure of the actual strength of the then proposed method (and given the accuracy with which differential cryptography computed 56 bits, I remain surprised they were able to provide that information, as the recommendation clearly communicated to a skilled adversary that they had developed and were leveraging this significant unpublished technique at that time).
- dogma1138 7y ago
  Differential cryptanalysis was discovered initially by IBM while working on DES. The NSA made design alterations to DES to make it more resistant to DCA. However they didn’t publish those design decisions because they would reveal the existence of the technique, which at the time was applicable to many other ciphers including nearly all Soviet ciphers. The NSA didn’t backdoor DES the key shortening and the addition of parity bits was an attempt to make it more resilient to DCA while making it only slightly more susceptible to brute force attacks. It’s also important to note that DES was a 1970’s cipher which survived long enough to be relevant, even a 1080ti would take 2 or more days to crack a single DES key and its considerably more powerful than likely all the super computers in the world combined in say 1983. The NSA isn’t your friend for sure but they know exactly what they are doing.

by mtgx 7y ago

yodon 7y ago

The NSA also didn't want to explain their reasoning for shortening the DES key size to 56 bits. It took many years for the public crypto community to discover the technique called differential cryptography, at which point it became clear that the actual theoretical maximum strength of the DES crypto system was 56 bits and the originally proposed longer key sizes didn't actually offer any true strength over 56 bit key variants. It remains important to be distrustful of all things in cryptography, but it also remains important that the one time the NSA had a huge and apparently negative impact on the strength of a crypto system, they were actually not weakening the system at all they were instead providing the public community with a significantly more accurate measure of the actual strength of the then proposed method (and given the accuracy with which differential cryptography computed 56 bits, I remain surprised they were able to provide that information, as the recommendation clearly communicated to a skilled adversary that they had developed and were leveraging this significant unpublished technique at that time).

dogma1138 7y ago
Differential cryptanalysis was discovered initially by IBM while working on DES. The NSA made design alterations to DES to make it more resistant to DCA. However they didn’t publish those design decisions because they would reveal the existence of the technique, which at the time was applicable to many other ciphers including nearly all Soviet ciphers. The NSA didn’t backdoor DES the key shortening and the addition of parity bits was an attempt to make it more resilient to DCA while making it only slightly more susceptible to brute force attacks. It’s also important to note that DES was a 1970’s cipher which survived long enough to be relevant, even a 1080ti would take 2 or more days to crack a single DES key and its considerably more powerful than likely all the super computers in the world combined in say 1983. The NSA isn’t your friend for sure but they know exactly what they are doing.