Who's responsible when AI goes wrong?

Disagree with it?