So as you may have heard, the Adobe user database with 130 million records was leaked recently. I want to find out what password I used for the Adobe account I created to make sure that I'm not using it anywhere else. But how do I decrypt the Triple DES ECB-mode encrypted password? Is there a way to do this without knowing the Adobe Cipher Key? Is there a way to get the Cipher Key?

  • zdnet.com/…
    – hunter
    Nov 13 '13 at 2:43
  • @hunter I want to know my password in plaintext. I can't remember it. I already know that my email is in the database.
    – user10379
    Nov 13 '13 at 5:44
  • The only easy way to obtain the key is from the server. The attackers might have done so, but AFAIK they didn't publish the key. So only thing you can do is check which users used a nearly identical password and look at their password hints. Nov 13 '13 at 8:32
  • This question appears to be off-topic because it is about how to work with the cryptography used by a particular breach of a particular service. It would be appropriate for Information Security. Nov 13 '13 at 13:24
  • Do we have single key in a server or multiple.?
    – Xsecure123
    Feb 25 '14 at 8:37

The encryption key was never published and triple DES itself is strong enough to not brute force, so you have to use some of the other mistakes to attempt to recover your password.

Your options are: exploit the fact that they used ECB and use the hints.

To do this you will find your encrypted password and see if other users had the same block. For example, say your encrypted password blocks were (shortened for simplicity): 87bef a31cd and you find another user that had 87bef 37cfe you know that the first 8 characters (since DES blocksize is 8 bytes) of your password are the same as the first 8 of the other user. Then you can use their hint to help figure your password out. Same process applies to subsequent blocks.

  • 1
    Could you elaborate on why 3DES is strong enough to not brute force? How long would it take?
    – Ehryk
    Nov 15 '13 at 2:33
  • What key length CAN be reasonably brute forced? Is there a practical formula, key length 80 ~= 3 months or something?
    – Ehryk
    Nov 15 '13 at 2:34
  • 1
    There is general consensus among cryptographers that anything with at least 80 bits of security is too much to bruteforce. To give a timing estimate would vary greatly depending on comptuation power. 3DES offers 112 bits of security (depending on configuration).
    – mikeazo
    Nov 15 '13 at 13:27
  • 1
    Keep in mind those bit lengths only apply to symmetric encryption (a government may be able to crack a 1024-bit RSA key today). Also, various algorithms have weaknesses, so it's not safe to assume that 80-bits is always going to be safe. Dec 20 '13 at 15:12
  • @Ehryk, distributed.net's effort to brute-force 72-bit RC5 is currently projected to exhaust the keyspace in 200 years. 168-bit 3DES is considered to provide the equivalent of 112 bits of security because of weaknesses in the construction.
    – Mark
    Feb 24 '14 at 1:36

A very good explanation can be found at this url: http://nakedsecurity.sophos.com/2013/11/04/anatomy-of-a-password-disaster-adobes-giant-sized-cryptographic-blunder/

The idea is exactly what mikeazo said in an other answer.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy