Re: Truly scary SSL 3.0 vuln to be revealed soon:

["Ben Lincoln (0E1C7DBB - OSS)" <0E1C7DBB () beneaththewaves net>]

On 2014-10-15 02:13, Pierre Schweitzer wrote:
> I've a naive question regarding the vulnerability, actually.
>
> It says you can recover plain text of ciphered text, using a specific
> method.
> But, in the end it means you'll have plain text + ciphered text of the
> same text. Does that mean you can easily bruteforce the key that was
> used? So that you can actually, if you logged the complete session,
> decipher the whole session of the user? And not only the cookie?
> Or breaking the key would be too complex yet?

Hi Pierre.

For modern block ciphers (e.g. AES, or even 3DES), known-plaintext 
attacks still generally require the entire keyspace to be brute-forced, 
which is not practical using the technology available today.

Think about the Adobe credential breach. There are many thousands of 
known plaintext + ciphertext pairs there (the same 3DES key was used to 
encrypt all of the passwords, and the passwords for many users were able 
to be recovered based on a combination of ECB-mode encryption + 
plaintext password hints), but the actual key was never recovered even 
with all of that data to work with.

- Ben