The use of autonomous weapons (AWS) has been hotly debated since their creation around 1900. Although research has been published on the principles of using AWS, little has been published on the relationship between AWS and just war theory (JWT). Moral Responsibility in Twenty-First-Century Warfare is a prophetic edited volume that explores the ethical dimensions of AWS and JWT. Published at the perfect time, the authors provide answers to questions that are just starting to appear in public policy and academia. Editors Steven Roach and Amy Eckert apply revisionist principles of JWT to modern warfare, in particular AWS.
At the heart of the book is the question of whether to overhaul JWT so that it can be applied to the current and future actions of AWS. The authors consider where the human responsibility gap starts and ends with AWS; they examine whether robots should be held accountable for their actions and if they have any rights. The frequent use of AWS in low-level conflict situations that blur the line between war and peace complicates any conception of JWT.
There are five themes in the book: “the deep morality of war, the reconfiguration of war ethics, the possible end of just war, a moral groundwork for robot rights and responsibilities, and the ethical uncertainty of advancing morality and accountability” (p. 4). What follows is a short synopsis of each contribution, with some critiques and suggestions where appropriate.
In chapter 1 Peter Sutch argues that legalist JWT is still applicable to twenty-first-century warfare and that international institutions are already capable of dealing with AWS human rights abuses, even if it is only a small group of states that commit them. Sutch contends that delegating authority to smaller entities to deal with these issues is unnecessary. Given that one of the greatest strengths of AWS is their speed, an increase in the regulation of AWS will decrease their efficacy by making them slower to deploy. This goes counter to Sutch’s argument.
In chapter 2 Laura Sjoberg presents a feminist critique of JWT. She states that this theory inherently contributes to violence because it accepts war as a possibility from the start, it does not have any foundation on which to evaluate war on a continuum, and it condones war as just. Sjoberg suggests that scholars personally question the privileges they are given to study concepts such as war, examine whom their work affects, create moral structures from the ground up to study violence, and examine different types of violence to further develop the field.
In chapter 3, Thomas Doyle compares and contrasts the moral responsibility for contemporary nuclear deterrence to that for AWS. Is nuclear deterrence ethical, and if so, does it steal moral agency from human beings? Does the use of AWS do the same? Doyle finds that human agency is lost to some extent in nuclear automatic processes and in AWS, even though there is some human control over AWS.
Sommer Mitchell in chapter 4 examines private military and security companies (PMSCs) as they help shape moral authority in twenty-first-century warfare. PMSCs have secured their legitimacy to use force through the Montreaux Document, the International Code of Conduct for Private Security Service Providers (ICoC), and the ICoC Association. However, Mitchell believes that these guidelines are geared to making profits for the industry and result in more violence because of their vague language. A code of conduct for JWT should include PMSCs and their personnel.
Chapter 4 provides a thorough analysis of the documents that justify and regulate PMSCs. However, what is missing in this analysis is any justification of governments’ use of PMSCs. Governments use these companies to do work that falls outside the confines of international law, to sidestep blame for actions that they commit, and to save billions of dollars while ignoring the long-term health care of PMSC personnel. PMSCs are here to stay as long as governments can use them as a cheap source of labor and as a scapegoat for their own actions.
In chapter 5, David Gunkel explores the rights of killer robots in a futuristic analysis that is probably much closer to our present reality than we realize. He examines the low status of war service animals under international law and claims that discussion of the rights of robots may lead to an increase in laws protecting service animals in battle. Instead of two classifications of human beings or tools, which currently include both animals and robots, Gunkel suggests additional categories that can be applied to robots and animals. Given that people attribute human characteristics to both animals and artificial intelligence (AI), additional classifications would serve human–robot teams that frequently use AI technology. It would have been interesting to read about proposed rights for animals and robots, had the author taken the analysis one step further.
Jai Galliott looks at the responsibility of human beings in JWT and how AWS have changed human responsibility in chapter 6. There is no one human being responsible for the deployment of AWS; instead, a huge system comprising many people develops, maintains, and deploys these weapons. Galliott advances the concept of institutional agents as a means to close the responsibility gap between humans and JWT.
In chapter 7, Jason Scholz, Dale Lambert, Robert Bolia, and Jai Gaillott make the case for AI weapons, stating that there is actually a time and place for killer robots to take human life. By designing robots with ethical constructs, the authors argue that AI can be quite useful. They develop a set of ethical guidelines for AI weapons and consider the competencies needed to make them. This chapter provides a deep discussion of how to create ethical killer robots, and the authors’ suggestions are highly innovative.
Overall, this book makes the case for AWS and suggests constructs to further regulate their use. In particular chapters 4 and 5 are must-reads. Much of the book is theory based, although there are valuable recommendations, particularly in chapter 7, for their use while still following the constructs of JWT.
Like any great book, there are a few areas in which it could have been stronger. Discussions of the practical applications of AWS using case studies such as the United States Air Force and of real-world scenarios where AWS were employed would have been helpful. Most scholars tend to oversimplify the processes of using AWS: collecting intelligence on a target, finding a target, and deciding to kill a target require the involvement of numerous political actors along the way and are relatively organized, cautious, and legal processes. No one person makes the decision to kill or target anything or anyone. In addition, AWS are most frequently used for reconnaissance and troop assistance.
In conclusion, Moral Responsibility in Twenty-First-Century Warfare is a valuable addition to the study of AWS and JWT that answers important ethical questions of modern warfare.