Beyond the Battlefield. The use of Autonomous Weapons System in War and The Crisis of Moral Accountability. An Analysis of the Limits Just War Theory Places on Combatants under jus in bello.
Onen Omony Kerali Godfrey
09 October 2025.
Abstract
Modern warfare has changed since 2012 following the invention and subsequent introduction and use of Autonomous Weapons Systems (AWS) in wars. This article confronts a central crisis in modern military ethics: the collapsing link between human agency and moral accountability in warfare. While the traditional principles of jus in bello—discrimination, proportionality, and, crucially, accountability—have long provided an ethical framework for combat, new technologies are rendering this framework obsolete. Through a critical analysis of Jonathan Parry’s work on the inherent tensions of moral consistency in war, and Robert Sparrow’s seminal “Killer Robots” thesis, this paper argues that Autonomous Weapons Systems (AWS) create an unbridgeable “responsibility gap.” It demonstrates that neither programmers, commanding officers, nor the machines themselves can satisfy the fundamental requirement for a morally culpable agent to be held responsible for wartime atrocities. The article concludes that the deployment of AWS is inherently unjust according to the tenets of just war theory, as it severs the essential chain of moral accountability, threatening to return warfare to a state of indiscriminate violence without answerability.
Key Concepts: Modern Warfare; Autonomous Weapons Systems; Moral Accountability;
Jus in Bello; Ethical Framework ; Killer Robots; Wartime Atrocities; Just War Theory.
Introduction
The longstanding tradition of Just War Theory doesn't end with the justification for going to war (jus ad bellum). It also provides a critical ethical framework for how war is fought: the jus in bello. These rules, governing the conduct of combatants, are meant to impose moral limits on the chaos and destruction of armed conflict. But how robust are these limits in the face of modern warfare? Drawing on the insights of Jonathan Parry's "War and Moral Consistency" and Robert Sparrow's "Killer Robots," this post will explore the intense pressure placed on the principles of jus in bello. We will examine Parry's argument that the very structure of war challenges the coherence of individual moral responsibility, and Sparrow's provocative claim that the development of autonomous weapons may render traditional ethical frameworks obsolete, creating a "responsibility gap" that could fundamentally undermine the project of restraining war's violence. This paper therefore examines the limits placed on the conduct of combatants fighting wars by just war theories, and an assessment of whether autonomous weapons systems (AWS) can comply with jus in bello principles. It examines the constraints of jus in bello in warfare and evaluates whether autonomous weapon systems (AWS) can satisfy these ethical standards. Jus in bello, the ethical framework governing the conduct of warfare, serves as a cornerstone of just war theory. These principles impose moral limits on combatants, aiming to minimize unnecessary suffering, safeguard non-combatants, and preserve the moral integrity of armed conflict. Central to jus in bello are three key tenets: discrimination (distinguishing combatants from civilians), proportionality (balancing military objectives against potential harm), and accountability (holding individuals responsible for wartime atrocities) (Parry 2020, pp. 692-694). By analysing these principles, particularly discrimination and proportionality, we can assess the ethical implications of AWS, as critically explored in Sparrow (2007, pp. 62–77).
The principle of discrimination mandates that combatants target only legitimate military objectives while avoiding harm to civilians. This distinction upholds the moral boundaries of warfare. For instance, deliberately attacking a civilian hospital constitutes a clear violation of jus in bello, whereas striking a military base may be permissible if it serves a strategic purpose. Similarly, the principle of proportionality requires that the anticipated military advantage of an action justifies the expected harm to civilians and civilian infrastructure. Even when targeting legitimate military objectives, the use of excessive force, such as deploying a nuclear weapon against a minor outpost, would violate this principle (Parry 2020, pp. 692-694). Sparrow (2007, p. 67) underscores that just warfare necessitates a morally and legally accountable agent for combat-related deaths as he notes "It is a fundamental condition of fighting a just war that someone may be held responsible for the deaths of enemies killed in the course of it. In particular, someone must be able to be held responsible for civilian deaths."(Sparrow 2007, p. 67). In traditional warfare, responsibility falls clearly on soldiers, commanders, or political leaders. However, AWS disrupt this framework by operating independently of direct human control, creating what Sparrow (2007, p. 68) identifies as the responsibility dilemma thereby aising the fundamental question of whether wars can be fought justly using autonomous weapons systems (AWS).
What are Autonomous Weapons Systems (AWS)?
Autonomous Weapons Systems (AWS), often described as "killer robots," mark a major shift in warfare. These systems can identify and engage targets without direct, real-time human control, unlike conventional weapons, where a human makes the final decision to use lethal force. Because autonomy exists on a spectrum, it is important to distinguish between different levels of independence in unmanned and robotic systems. The U.S. Department of Defence identifies four categories: human-operated systems, systems with delegated independent functions, systems that operate independently under human supervision, and fully autonomous systems that, once activated and programmed, require no further human involvement. A system’s autonomy depends largely on its computational capacity. Algorithms enable it to process information and perform tasks independently, while hardware converts those calculations into physical action. It is also important to distinguish autonomous systems from automatic ones. Autonomous systems can act in unfamiliar environments without human input, whereas automatic systems follow pre-set controls without direct human operation. Automatic weapons such as rapid-fire systems and landmines have long existed, but they differ from autonomous systems in decision-making capacity. For example, a landmine is triggered by pressure, whereas an autonomous weapon may fire based on selective indicators such as heat or motion (Antebi 2019, pp. 78-79). Many current weapons are automated rather than autonomous, performing limited functions such as guiding a missile to a chosen target. A true AWS, however, shifts humans from being "in the loop" to "on the loop" or even completely "out of the loop." In its most advanced form, an AWS could independently search an area, identify targets, decide which to attack, and carry out the strike. Its autonomy can vary across functions such as movement, targeting, and engagement, and may apply in both split-second defensive actions and longer missions. This development raises serious ethical and legal concerns, especially in relation to jus in bello. A central issue is the "responsibility gap": if a fully autonomous weapon unlawfully kills a civilian, it is unclear who should be held accountable. The responsibility cannot easily rest with the soldier who made no direct decision, the commander who issued only a general order, the programmer who could not predict every scenario, or the machine itself. This gap threatens the very possibility of moral accountability. AWS also raise doubts about whether machines can comply with core principles of International Humanitarian Law, especially distinction and proportionality. These principles require careful human judgement to distinguish combatants from civilians and to weigh military advantage against civilian harm—tasks that critics argue exceed the capabilities of algorithms. Ultimately, AWS are more than a new military technology; they may transform the nature of war by weakening the link between human agency and lethal force (Harwood 2024, pp. 1-2).
Can autonomous weapons systems be used justly in war? A reference to Robert Sparrow’s “Killer Robots”
Sparrow's article Killer Robots (2007, pp. 62-77) presents a critical examination of whether autonomous weapons systems (AWS) can satisfy jus in bello principles, with particular focus on the requirement of accountability. His central argument identifies a fundamental responsibility gap created by AWS, rendering just attribution of blame for war crimes committed by these machines impossible. He delivers a definitive rejection of AWS compliance with just war standards, systematically analyzing and dismissing the three potential candidates for assuming responsibility for AWS-related war crimes: the programmers who create these systems, the commanding officers who deploy them, and the weapons systems themselves. Through this critical analysis, he demonstrates that none of these parties can adequately fulfill jus in bello's accountability requirements, thereby creating what he terms the " responsibility gap" dilemma (Sparrow 2007, p. 68).
Responsibility of AWS Programmers.
One might argue that AWS programmers should bear responsibility for their machines' actions, as war crimes committed by autonomous systems could result from programming flaws or negligence. However, Sparrow (2007) systematically refutes this position through two key arguments. First, he establishes that when manufacturers explicitly warn military forces about a system's potential targeting errors, accountability for resulting war crimes transfers to those who deploy the AWS despite these known risks (p.69). Second, he demonstrates how the autonomous nature of these systems fundamentally disrupts traditional chains of responsibility. As Sparrow notes: "The more the system is autonomous, the more it has the capacity to make choices other than those predicted by its programmers". This inherent unpredictability means that beyond a certain threshold of autonomy, AWS's actions become untraceable to its original creators. He illustrates this point through the analogy: ”holding programmers accountable for their autonomous creations would be as unreasonable as prosecuting parents for crimes committed by their adult children (Sparrow 2007, p.70).
The Commanding Officer’s Responsibility
A fundamental question arises whether military doctrine holding commanders accountable for subordinates' actions could extend to autonomous weapons systems (AWS). Sparrow acknowledges this principle applies to predictable conventional weapons like artillery, where off-target strikes represent foreseeable risks. However, he argues this framework fails for AWS precisely because they are marketed as intelligent systems capable of independent decision-making. When AWS commit war crimse or strike unintended targets, holding commanding officers responsible becomes unjustifiable. As Sparrow explains: "The more autonomous these weapons are, the larger [the] risk [of unfair blame] looms. At some point, it will no longer be fair to hold the Commanding Officer responsible for the actions of the machine" (Sparrow 2007,p.70). This reasoning demonstrates that punishing humans for machines' independent decisions violates core principles of moral agency (Sparrow 2007, pp. 70-71).
Can AWS be held accountable for war crimes?
The possibility of assigning moral responsibility to AWS themselves presents significant philosophical challenges. Sparrow examines whether even the most advanced artificial intelligence could be considered morally responsible and thus subject to punishment for war crimes, ultimately concluding this is impossible for two fundamental reasons. First, meaningful punishment requires the capacity for suffering—the offender must be able to experience its consequences. However, AWS lack consciousness, emotions, or any sense of guilt that would make punishment morally significant. As Sparrow asserts: "In order for our treatment of the machine to count as punishment, it must be capable of suffering in ways that might motivate the same set of responses that we have as a matter of course to human beings" (Sparrow 2007, p. 72). Second, AWS cannot achieve true moral personhood. Even if they replicate human-like reasoning, they lack the moral depth and genuine understanding required for accountability. Punishing an autonomous weapon for war crimes would be as logically incoherent as prosecuting a gun for firing a bullet. Sparrow’s argument extends beyond AWS to warn against any technology that undermines human accountability in warfare, emphasizing the dangers of delegating lethal decisions to autonomous systems (Sparrow 2007, pp.71–72).
The Child Soldier Analogy
Sparrow (2007, pp. 73-74) employs the analogy of child soldiers to illuminate the ethical ambiguity surrounding AWS. He argues that both child soldiers and AWS occupy a problematic middle ground in moral responsibility - capable of making battlefield decisions yet lacking full moral comprehension. Child soldiers, due to their cognitive immaturity, cannot fully grasp the ethical gravity of killing in war and thus cannot bear complete responsibility for their actions. Similarly, commanders cannot be held fully accountable for child soldiers' conduct, as children act with unpredictable autonomy unlike adult soldiers under clear command structures (Sparrow 2007, pp.73-74). AWS exist in this same ethical "grey area" - autonomous enough to disrupt traditional chains of responsibility yet insufficiently autonomous to assume moral blame for war crimes. This creates what Sparrow identifies as a responsibility gap: atrocities may occur without any morally culpable agent. Just as child armies operate without meaningful oversight, leading to indiscriminate violence, AWS undermine the fundamental principles of just war theory by removing human moral agency from lethal decision-making.
Conclusion
As states and non-state actors increasingly use autonomous weapon systems (AWS), Sparrow argues that deploying AWS is inherently unethical as it undermines jus in bello's principle of moral responsibility. Without a culpable agent, warfare risks devolving into indiscriminate violence, where deaths are neither justified nor accounted for. Holding commanders responsible for machine-driven actions they cannot control is unjust, as is treating enemy combatants’ lives as expendable with no one answerable for their deaths (Sparrow [insert page numbers if available]). Sparrow’s analysis underscores the ethical dilemma posed by AWS. Deploying AWS cannot be justified, since they are inherently unfair—not only to civilians living in conflict zones, but also to military personnel who remain legally responsible for the actions of AWS. The unpredictability of AWS creates an unacceptable accountability gap: neither programmers nor operators can fully control outcomes, yet commanders remain legally exposed. Current international law and military protocols lack the framework to address this challenge. At its core, Sparrow’s argument rests on jus in bello’s demand for accountability. A just war cannot incorporate AWS without violating this principle. The proliferation of autonomous weapons threatens to erode the ethical foundations of warfare. A legally binding international ban on fully autonomous weapons systems is not merely prudent but morally imperative to protect combatants and civilians alike and preserve the ethics of armed conflict against the atrocities and erosion that autonomous killing machines represent.
References
Antebi, L., 2019. The proliferation of autonomous weapons systems: Effects on international relations. National Security in a “Liquid” World, pp.75-92.
Harwood, S 2024. A cybersystemic view of autonomous weapon systems (AWS).
Technological Forecasting and Social Change, Volume 205, 123514.
Parry, J. (2020). War and moral consistency.
Sparrow, R. (2007). Killer robots. Journal of applied philosophy, 24(1), 62-77.
Comments