Machine Ethics

Machine Ethics or AI Alignment? compares moral-theory implementation and artificial moral reasoning with alignment to human values. The fields overlap in their concern for behavior, but disagree or vary in how targets are justified.^[1]

Operational success requires both a defensible norm and reliable implementation. Human consensus alone need not establish moral correctness; a formal norm alone need not ensure safe conduct.

Responsibility around the machine

Meaningful Human Control adds a question about the people who design and use an autonomous system: can they understand its limits and recognize their responsibility? Santoni de Sio and van den Hoven distinguish this from merely giving a person an approval role. Their philosophical account also allows human control over wrongful objectives, so control does not resolve the justification of the norm being implemented.^[2]