When it comes to ethics, artificial intelligence is just intelligence, not a person with rights and responsibilities.
It's owner is the person in which custody these machines live in and they should be fully responsible for actions of their machines.
Even if they act human, they are just machines owned by someone and they should be treated as such.
That's fair. The reason I ask this question is because Europe is now deciding whether or not to grant robots "personhood". If this becomes the new norm, we will have a lot more to think about and it will take the idea of "ethics" to a whole new level.
https://slate.com/technology/2018/04/the-eu-is-trying-to-decide-whether-to-grant-robots-personhood.html