Encrypting same data with many keys...

[nospam-seesignature at ceddec.com]

On Tue, 12 Aug 1997, Ray Arachelian wrote:

> A known plaintext attack won't help you to break the keys unless you have
> one of the eight keys, but will having many keys that encrypt the same
> data significanltly weaken the security of that tiny chunk of data?
> 
> And no, I don't mean, there's N keys so the odds of brute forcing the data
> is now N times easier.  Assume we're using 128 bit Blowfish/Idea or
> better, and discarding weak keys.  Are there any differential or other
> cryptanalysis methods to use the eight resulting cyphertexts to get at the
> data other than brute forcing it if you don't know any of the keys?
> 
> What if instead of using a private key cypher, we used a public key
> cypher?  Would that make any difference in attack methods?

The only thing I can think of is if you use something like CFB mode, and
the IV is also the same at the beginning, the first 8 bytes will leave a
hole - I don't remember exactly, but I was burned by exactly this when I
saw 8 bytes of plaintext after resetting the IV in an app that xors some
encrypted blocks of data to do something else.

A PK to encode the conventional key works better since you can do a long
or complex conventional key and other material such as an IV once, and
then bury that several times.

--- reply to tzeruch - at - ceddec - dot - com ---