Seems like a good idea, but how would it be implemented in a way that prevents the AI crawlers from hoovering up the data anyway?