By most implementations of copyright laws, your embodiment of B was derived from X, and publishing it without permission is an infringement. If you had first generated B and then derived A from it using X then your publication of A would instead be an infringement.
Nothing inherent in the bit strings A and B make them infringing: it is their provenance that makes publication infringing. The fact that a random looking A and B can be combined to form X is very strong evidence that one or the other was derived from X and therefore publication without permission is infringing.
A court may find it difficult to determine which of you or your friend actually committed the infringement, but they can find that on the balance of evidence that at least one of you was, and hold you jointly and severally liable.
If you are actually an innocent party, you might use your earlier publication as evidence that the other party committed the infringement. After all, if you first published A and then someone else publishes B that xors with A to form X, then it’s much less likely that you used their innocent B to derive A instead of the reverse.
By most implementations of copyright laws, your embodiment of B was derived from X, and publishing it without permission is an infringement. If you had first generated B and then derived A from it using X then your publication of A would instead be an infringement.
Nothing inherent in the bit strings A and B make them infringing: it is their provenance that makes publication infringing. The fact that a random looking A and B can be combined to form X is very strong evidence that one or the other was derived from X and therefore publication without permission is infringing.
A court may find it difficult to determine which of you or your friend actually committed the infringement, but they can find that on the balance of evidence that at least one of you was, and hold you jointly and severally liable.
If you are actually an innocent party, you might use your earlier publication as evidence that the other party committed the infringement. After all, if you first published A and then someone else publishes B that xors with A to form X, then it’s much less likely that you used their innocent B to derive A instead of the reverse.