Wikileaks' insurance.aes256 file and AES Crypt

[paulej]

We recently saw a huge spike in traffic to the AES Crypt web site and the number of downloads of the program increased significantly. While we're certainly happy to see people take an interest in AES encryption technology and AES Crypt, in particular, what we discovered as we tried to find out why there was a sudden and significant increase was that Wikileaks had apparently published an encrypted file called "insurance.aes256". You can read more about it here.

Naturally, people from all over the world started trying to figure out how to decrypt the file and went searching for the tools to do it. Curious as to whether AES Crypt was used to encrypt the file, we inspected the file and determined that AES Crypt was not used. Rather, it appears to have been encrypted with OpenSSL. While AES Crypt and OpenSSL both support 256-bit AES encryption, the file formats are not compatible. As such, it is impossible to decrypt the insurance.aes256 file using AES Crypt.

As the article from France24 points out, trying to decrypt that file by guessing is really a lost cause. AES is a very strong encryption algorithm that, as far as the world knows, has no weaknesses. So, without the password, it is likely going to prove impossible to read the file with OpenSSL. I just wonder how many of the world's computers have been put into service to crack the password?

[Sergej.Kaempf]

Interesting, but maybe the file contains just some random data.

Is the AES file format, which is described on your side, a standard or a design by the AES Crypt-Team? Because it must have some benefits over the OpenSSL file format (raw data?), otherwise it wouldn't make sense to design a new one.

[paulej]

The "insurance.aes256" may very well contain random data. There's no way for me to tell.

As for the file format, there is no standard file format for AES-encrypted files. OpenSSL has its own, WinZIP has its own, and we introduced our own. Since AES Crypt is focused strictly on encrypting files using AES, we wanted a format suitable for the purpose. As we designed AES Crypt, we had a few requirements, including the ability to detect whether the file had been manipulated (we use an HMAC for that), a signature that we can use to detect that a file is generated by AES Crypt (we insert 'AES' at the start of the file), etc. More recent versions introduce the ability to insert unencrypted metadata into the file (e.g., the program used to create the .aes file) and to more rapidly detect an incorrect password.

The .aes file format we use is published here:
http://www.aescrypt.com/aes_file_format.html

[Sergej.Kaempf]

Too bad, there is no standard file format for AES, but i can live with that, especially when AES Crypt is easier to use, if you don't want to work with a console. And maybe in future OpenSSL will adapt your file format or vice versa, because the tags-system seem useful, as long you don't give away too much information.

The random data: I just tought it would be funny, how they publish a file which contains no information at all to keep the governments busy. And the goverments must know the things they have to hide, so the only reason to decrypt it, would be to see, how much do WikiLeaks really know. And random data would be perfect to bring them sleepless nigthts.

Who knows how much of WikiLeaks is true or not, but they should carefull, especially in russia:
http://en.wikipedia.org/wiki/List_of_jo ... _in_Russia.

[deewany]

What happened to all the wikileaks articles not related to the Iraq or Afghan war? The wikileaks website use to have articles on everything but now there is only articles on the Iraq or Afghan war. Whats going on with their website?