You write as if Asimov's fictional three laws were already the basis for AI systems and then give examples (some equally fictional) where they have not been applied. Given your premise is wrong, any conclusion that you draw from it is at best unproven.