Robots and Respect: Assessing the Case Against Autonomous Weapon Systems

The prospect of “killer robots” may sound like science fiction. However, the attention given to the operations of remotely piloted drones in recent years has also spotlighted the amount of research that is being conducted on weaponized robots that can select and attack targets without the direct oversight of a human operator. Where the armed services of the major industrialized countries were once quick to distance themselves from the use of autonomous weapons, there is increasing speculation within military and policy circles—and within the U.S. military in particular—that the future of armed conflict is likely to include extensive deployment of autonomous weapon systems (AWS). A critical 2012 report on AWS by Human Rights Watch and the 2013 launch of an NGO-led campaign for a treaty prohibiting their development and use has intensified the ongoing ethical debate about them.

My aim in this article is twofold. First, I will argue that the ethical case for allowing autonomous targeting, at least in specific restricted domains, is stronger than critics have typically acknowledged. Second, I will attempt to defend the intuition that, even if this is so, there is something ethically problematic about such targeting. Given the extent of my ambitions, the dialectic that follows is somewhat complicated and for this reason it will be useful to briefly sketch an outline of the argument here.

My argument proceeds in three parts. In the first section I introduce a working definition of “autonomous” weapons and describe the military dynamics driving the development of these systems. In the second section I survey and evaluate the existing literature on the ethics of AWS. The bulk of this discussion is framed as an account of two “rounds” of debate between an influential advocate for AWS, Ron Arkin, and his critics. In the third and final section I turn to a deeper investigation of the philosophical foundations of the just war doctrine of jus in bello in order to develop a new account of the origins and force of the intuition that the use of killer robots would necessarily be morally problematic. I conclude that although the theoretical foundations of the idea that AWS are weapons that are evil in themselves are weaker than critics have sometimes maintained, they are nonetheless strong enough to support the demand for a prohibition of the development and deployment of such weapons.

To read or purchase the full text of this article, click here.