>The SHA-1 pre-image resistance would prevent that. No, I'm talking about computing something much less ambitious than solving pre-image resistance. Consider

>The SHA-1 pre-image resistance would prevent that. No, I'm talking about computing something much less ambitious than solving pre-image resistance. Consider the example SHA-1 hash[1]: SHA1("") gives hexadecimal: da39a3ee5e6b4b0d3255bfef95601890afd80709 gives Base64 binary to ASCII text encoding: 2jmj7l5rSw0yVb/vlWAYkK/YBwk= Embedded in that hex is "95601890" ... which could be somebody's driver's license number. Now that the number is publicly visible via some legitimate purpose other than illegally publishing drivers license numbers, we can then argue "How can a SHA-1 of a empty string illegal?!?" It's a similar concept to creating "vanity" public hashes for Tor or bitcoin addresses or finding a hash that has a certain number of leading zeros in "proof of work". Embedding arbitrary illegal information that way is more computationally feasible with brute force than perfect pre-image attacks. The underlying idea of the hack is the same : Find some way to make illegal data be interpreted in a totally different way that _is_ legal. It's very hard to do for large amounts of information such as 9 GB mp4 rip of a Blu-ray disc -- but easier for small amounts of information... such as the DeCSS C source code converted to an "interesting" prime number. [1] https://en.wikipedia.org/wiki/SHA-1#Example_hashes https://en.wikipedia.org/wiki/SHA-1#Example_hashes