To Start, If we follow the Assimov Robot Laws:
1) A Robot will not Harm or Allow a Human Being to be harmed.
2) A Robot will Follow the Directions given to it by a Human, as long as it does not conflict with the 1st Law.
3) A Robot will Protect itself to Survive, as long as it does not conflict with the 1st or 2nd Law.
But, it needs to be noted, that these Laws mean that :
- the Robot AI needs to be Aware of what a Human Being is,
- and be Aware of what Self represents to itself,
for the sake of these Laws being clearly and Absolutely enforced by the Robot AI.