The issue here is a deep one. Our current legal paradigm is skewed exclusively towards Imperative programming. The public and judges and most lawyers understand no other. We try to prohibit side-effects that the legal system doesn't want by limits to functions that might compute them. The net result, since side-effects of functions are quite unpredictable, is the only way to enforce undesirable side-effects is to censor the functions in advance. This amounts to pre-censorship rather than the present unsuccessful post-censorship. It also drives attempts at defining, in advance, which functions are off-limits conceptually (by software patents).
This amounts to taking away the right to program an (approximation to) a universal computing device. To the extent this approach succeeds, it will be to the detriment of computational freedom generally.
The line of defense in the long run *must* be the freedom to compute anything. Computation is like thinking. You can't censor it in advance by rules.