The US Department of Defense has revised its definition of artificial intelligence to grant tanks the ability to determine their own targets. Although they do not have plans to use AI instead of humans in the near future, they are reportedly making developments aimed at enhancing the capabilities of human soldiers. For example, an AI-equipped system today is capable of acquiring targets 3 times faster than a human. The trigger, for now, remains in the hands of humans. As far as I understand it; could those who say today, “I can have a nuclear bomb and I use it for the peace of humanity, but you shouldn’t have one,” say in the not-too-distant future, “My AI is not a killer, your soldiers made a mistake”? ( Source )