unpredictably dangerous rogue

No one's judging robots just yet. But would we judge one fully-tooled up with artificial intelligence and able to develop its own responses? Probably not. We'd say, No, it's acting according to its complex programmed logic, however refined that's become. And that's on us. Wouldn't stop us terminating it as an unpredictably dangerous rogue element in society. But we wouldn't actually apply moral judgement to it. We'd continue to reserve that for human behaviour (over a certain age and with observable functionality). Even though human behaviour is probably not that different.