There is a victim in those two analogies. Contrary to what one may think, not being able to generate hateful or "unsafe" text with AI does not make one a victim.