Xtykutl

I have written some code using the pycrypt library and I must be doing something wrong, but I can't figure out what it is that I am doing wrong. I can (nearly) decrypt messages with the wrong initialization vector even though I believe I am following their examples.

from Crypto.Cipher import AES

import os

from string import ascii_letters



key, iv = os.urandom(32), os.urandom(16)

cipher = AES.new(key, AES.MODE_CBC, iv)

plaintext = ascii_letters

plaintext += ' ' * (-len(plaintext) % 16)  # Padding

ciphertext = cipher.encrypt(plaintext)



cipher = AES.new(key, AES.MODE_CBC, os.urandom(16))

text = cipher.decrypt(ciphertext)

# text[16:] = b'qrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ            '

Now aside from the first 16 bytes of the decrypted text, you have everything decrypted correctly with a totally random choice of IV. Can someone help me figure out where I am going wrong?

That's normal for cipher block chaining. In CBC decryption, the IV is only necessary to reconstruct the first block of plaintext. The computation of other blocks of plaintext doesn't actually involve the IV. Here's a diagram (source: Wikimedia user WhiteTimberwolf, public domain):