very interesting, but can the one who made the "copy" claim that he just randomly generated the bytes, and didn't expect that it would end up as an encoded copy of the original?

That'd be perjury.

@B''HBi'ezras--BoruchHashem but made it clear for you, who is responsible.

Just another point though, given that it's purjery and that one party is definitely liable and lying, but how can the court determine which party it is? (See new comments to question)

@B''HBi'ezras--BoruchHashem That's a different question.

but technically that is what the body of the question asked

If an encrypted file can be decrypted and the original file recovered, the chance that the encrypted version was created by a random process is just the same as the chance that a random process would have reproduced the original video in full. Too small to consider. How long would it take to make a hi-res Mona-Lisa by throwing darts? How long for typing monkeys to reproduce Hamlet? No judge or jury is likely to believe that defense.

If the algorithm is not reversible, the output is not a copy. A hash, for example, is not a copy of the file being hashed.

@DavidSiegel a derivative work is still a derivative work even if it is not possible to recreate the original work from the derivative work. But a hash is not a creative work, so probably not copyright protected. A creative application of a blurring algorithm to some image, for example, may be both irreversible and a derivative work.

@phoog True. What I said is that the output of an irreversible process is not a copy. Note that a new copyright depends on originality not creativity. An unoriginal and uncreative derivative work may still be infringement, but a work so modified that no trace of the original can be discerned would not normally be treated as a derivitive work,

@phoog expounded on that little difference.

For example if one takes the text of a creative work, and XORs it with a truly random byte-stream, the result is also a random string, and no meaningful trace of the original remains (if the random input is not preserved). Such an output would not be a derivative work, even though it depends on the input work at every point. More reasonably, artists have started with an image but so mixed and transformed it that no visible trace of the original remains. In such a case the output will not usually be considered derivative.

If one makes a hash of a protected work, to serve as a unique identifier of that work, I suspect that hash would either not be treated as a derivative at all, or would be considered to be a fair use. I don't know of a case on-point for this, however.

@DavidSiegel a hash for identification would be most likely fair use in my eyes, kind of a checksum. "The original Cinema Cut has Checksum A, the Director's cut has Checkshum B,...." seems like a usage that would not infringe the market ever.

@B''HBi'ezras--BoruchHashem If the encoding scheme does not rely on a secret key, then one could simply decode your supposed "random" artwork and demonstrate that, when your artwork is decoded, it produces an output that is identical to a copyrighted work. If the encoding scheme relies on a secret key (i.e. an encryption algorithm), then one would need the key as well to demonstrate that. In some situations I believe you can be legally compelled to divulge a secret key that you know as part of an investigation.

@DavidSiegel if no judge or jury would believe it was randomlygenerated, then why do government schools teach that everything in the world came about through random chance, if even a short little movie can't?

@Brandin but thats the wole point of the question, what if the "secret" key is in the decoder program, or separately published in a different context than both the program and the encoded image? Can both or either the program writer and the encoder claim that there's no secret key, and whoever published the secret key is the one deliberately causing the work to be "decoded" into copyright content?

@B''HBi'ezras--BoruchHashem They can subponea the generator of the work to hand over the PC for inspection. If they find the tool deleted or missing, they can infer that it did exactly what the suing party alleges: turns the work into the other format. Only if they find the tool and it doesn't generate that other file with the input, they are exonerated.

@B''HBi'ezras--BoruchHashem Everything in the world did not come about through random chance. See biology.stackexchange.com/questions/tagged/evolution

@Brandin i agrree, it was carefully designed by the Creator through His speech and Will, which we can relate o with video encoding/encryptian

@B''HBi'ezras--BoruchHashem christianity.stackexchange.com

@Brandin i don belive i religion. the question was that courts shoudlnt accept the argument that it randomly came together, but the point is is that is exactly what the government teaches anyways, so it should be a valid argument