I disagree.

AI is a tool, and as such has been programmed. Can it be programmed for bad purposes, aka evil purposes, of course it can. An AI could possibly be programmed to tell people to commit all sorts of evil atrocities. But does that make the AI evil, does that make it satanic?

Let's suppose that AI was programmed to solve world hunger, the homeless problem, the miscommunication between nations, would that make it "saintly?" Again, we are imbuing human intention upon something that either is only doing what it is told to do, or it is teaching itself. Does either option mean it is "good or evil?"d