U.S. Defense Secretary Ashton Carter says the army will never give its robots “true autonomy” to kill, and while at first this seems like an eminently reasonable way to prevent a killbot apocalypse, the reality of the issue is a little more complicated.
“In many cases, and certainly whenever it comes to the application of force, there will never be true autonomy, because there’ll be human beings [in the loop],” Carter explained during a Thursday interview with two reporters from Breaking Defense.
That said, the U.S. Army is already “making big investments” in the technology and it’s working on getting autonomous robots to the battlefield. But, there’s “always going to human judgment and discretion.”
“That’s both necessary and appropriate,” Carter added.
The Pentagon has made statements to the same effect, but Breaking Defense pushed Carter on the possible downside to keeping humans in the loop. One of the reporters asked if such a hardline stance was akin “unilateral disarmament,” since there are serious concerns that China and Russia are exploring autonomous killer robots.
Fighting robots that need to wait for a human to give the OK could be at a disadvantage against A.I. that can shoot whenever it sees fit. Could ethical concerns and/or fear of Killbots running amok put the U.S. at a tactical disadvantage?
Carter seemed not to think so. He repeated his position, elaborating on what, exactly, the human role was going to be in a robot-human military alliance. Regardless of what type of force you’re working with, he explained that you “set things up [in advance], give orders and instructions such that everything that is done, is done in a way that is compatible with the laws of armed conflict… as well as American military doctrine.” That’s the plan for killer robots, too.
Even if the military doesn’t plan on creating truly autonomous killer robots, Carter says it will deploy them for a variety of other roles.
“People tend to want to think of autonomous systems for the use of lethal force,” Carter told Breaking Defense, “but their most likely applications in the near-term and mid-term are for such tasks as scanning networks for vulnerabilities, scanning incoming traffic, and doing the kind of work that a cyber defense analyst needs to do today by hand.”