3 ms·

This just exemplifies the NSA's role of being the crack dealer of information security. There's always a free sample, and it's always addictive enough to get yo

This just exemplifies the NSA's role of being the crack dealer of information security. There's always a free sample, and it's always addictive enough to get you coming back for more, at their costs. Google et al got info on actual threats, so when the NSA came back later to ask for "favors", a lot of critical analysis of what the actions really entailed that should have been done was left by the wayside. Then again, the NSA has a pretty long history of doing this. Look at the work the NSA did on DES, for example. They strengthened the S-boxes against differential cryptanalysis at the cost of ensuring that they'd be able to brute force things secured with the algorithm years before anyone else could.

pilsetnieks 12y ago
Not to be an apologist but I've only ever heard the first part - that they strengthened the S-boxes against differential cryptanalysis at a time when everyone else was still 10 years away from discovering it. If you mean with the second part that the key was shortened from 64 to 56 bits then that made it easier for everyone to brute force it, provided you had the resources.
- Sanddancer 12y ago
  Yes, it made it easier for everyone, however, because the NSA had a lot more money to throw at the problem, they could do such years ahead of anyone else. Also, the NSA originally was trying to get IBM to use a 48-bit key; 56 bits was a compromise.
  - csandreasen 12y ago
    The Technical Director of the NSA Information Assurance Directorate gave a less sinister explanation for this in a keynote talk[1]. It was dropped from 64 bits to 56 bits for two reasons: 1) they wanted to add 8 parity bits to make it more robust for tactical military use, and 2) they decided that 56 bits would be an acceptable key length to only last for a couple decades before it would have to be replaced. They had no way of knowing in 1976 what advances in cryptanalysis would be made over the next several decades, and wanted to make sure that the public wouldn't be tempted to rely on any one particular algorithm longer than was necessary. This might seem counter-intuitive at first, but remember that the biggest threat to NSA cryptography was from Soviet cryptanalytic work, not from the public sector. Public researchers would publish any advances they made. If the Soviets cracked DES there wouldn't be any public notification - they'd just siphon off sensitive data for years until they were caught. [1] http://vimeo.com/97891042 http://vimeo.com/97891042 (the whole talk is interesting, but jump to 22:50 for the DES part)

pilsetnieks 12y ago

Not to be an apologist but I've only ever heard the first part - that they strengthened the S-boxes against differential cryptanalysis at a time when everyone else was still 10 years away from discovering it. If you mean with the second part that the key was shortened from 64 to 56 bits then that made it easier for everyone to brute force it, provided you had the resources.

Sanddancer 12y ago
Yes, it made it easier for everyone, however, because the NSA had a lot more money to throw at the problem, they could do such years ahead of anyone else. Also, the NSA originally was trying to get IBM to use a 48-bit key; 56 bits was a compromise.
- csandreasen 12y ago
  The Technical Director of the NSA Information Assurance Directorate gave a less sinister explanation for this in a keynote talk[1]. It was dropped from 64 bits to 56 bits for two reasons: 1) they wanted to add 8 parity bits to make it more robust for tactical military use, and 2) they decided that 56 bits would be an acceptable key length to only last for a couple decades before it would have to be replaced. They had no way of knowing in 1976 what advances in cryptanalysis would be made over the next several decades, and wanted to make sure that the public wouldn't be tempted to rely on any one particular algorithm longer than was necessary. This might seem counter-intuitive at first, but remember that the biggest threat to NSA cryptography was from Soviet cryptanalytic work, not from the public sector. Public researchers would publish any advances they made. If the Soviets cracked DES there wouldn't be any public notification - they'd just siphon off sensitive data for years until they were caught. [1] http://vimeo.com/97891042 http://vimeo.com/97891042 (the whole talk is interesting, but jump to 22:50 for the DES part)