Peter Backes

"It's very low probability to have same 64-bit pseudo-random IV for same user, at least for Windows version of 7-Zip." If the same user encrypts many files, it becomes likely that he will hit the problem. If a lot of different users each encrypt some files, it becomes likely that one of them will hit the problem. The issue is thus far from being theoretical. Please learn your math. The security claim for AES128 is 128 bits of security. By using only 64 bits of IV, you are significantly weakening...