ContentID stores some kind of hash to identify music; isn't this "hash" just another version of the same work and therefore subject to copyright?
My point is it's likely that ContentID needs some kind of permission from the artist to even work, because in order to identify someone's music it needs to store that music (or some representation thereof), and if it does that without the artist's permission, it's infringement.
So one option could be to get out of ContentID entirely and then forbid Google to go after her fans.
It seems fairly clear to me from the article that this is entirely her choice already. She has three options: sign the new agreement and continue receiving cash for videos including her content, drop out of the agreement and stop receiving cash but leave the videos up, or drop out of the agreement and take all the videos down. She has full control over this.
Nope. A SHA-256 cryptographic hash isn't a "version" of a file in any sense of the word. You couldn't (for example) run a binary hash of an executable, nor can you convert it into an executable file by compiling it or reversing the hash algorithm.
My point is it's likely that ContentID needs some kind of permission from the artist to even work, because in order to identify someone's music it needs to store that music (or some representation thereof), and if it does that without the artist's permission, it's infringement.
So one option could be to get out of ContentID entirely and then forbid Google to go after her fans.