True, I usually wait until the facts are in, but I'm more curious about the question itself. At what point do semi-autonomous programs or constructs become liable for their words or actions.