hridal.

Would autonomous battlefield robots be a good thing?

12 January 2026

“Building autonomous battlefield robots would mean no more allied soldiers would die needlessly, and so would be a good thing.” Discuss.

The prospect of autonomous battlefield robots is often presented as a moral advance in warfare. If machines could replace human soldiers on the battlefield, then allied personnel might be spared death, injury, as well as psychological trauma. On this basis, it is frequently claimed that autonomous battlefield robots would be a good thing simply because they prevent allied soldiers from dying needlessly. At first glance, this reasoning appears compelling, particularly in a context where war is already understood as morally tragic.

However, the move from fewer allied deaths to the conclusion that autonomous battlefield robots are therefore morally good involves a significant logical step. Preventing harm is a morally relevant consideration, but it does not automatically settle questions of moral permissibility. The claim assumes, without argument, that reducing allied casualties is sufficient to justify how combat is exercised. Whether this assumption can be defended is precisely what is at issue.

Clarifying the debate requires first defining what is meant by autonomous battlefield robots. In this essay, they will be accepted as weapons systems capable of selecting and engaging targets without direct human intervention once deployed. This definition follows the characterisation used in academic and policy discussions of lethal autonomous weapons systems, particularly by Human Rights Watch and the International Committee of the Red Cross, which emphasise the absence of meaningful human control over individual targeting decisions (Docherty, 2025).

With this definition in place, this essay rejects the claim that autonomous battlefield robots would be a good thing simply because they prevent allied soldiers from dying needlessly. While harm reduction counts in favour of autonomous systems, it is not morally sufficient. By evaluating the strongest defences of the claim and weighing them against objections grounded in moral responsibility, moral equality, and the transformation of war, this essay will argue that the justification offered by the claim ultimately fails.

The most powerful argument in favour of autonomous battlefield robots appeals to harm reduction. At a deeper level, the harm-reduction argument illustrates a broader problem in the reasoning of the claim. Even if autonomous battlefield robots prevent allied soldiers from dying, it does not follow that their use is therefore morally good. War inevitably involves pain and suffering, so if autonomous systems could replace human soldiers, military objectives might be pursued without exposing allied personnel to lethal risk. Fewer soldiers would die, fewer families would grieve, and fewer individuals would be required to make fatal decisions under extreme stress. This argument fits naturally within a utilitarian framework. Utilitarianism evaluates actions by their

consequences, aiming to maximise overall wellbeing or minimise suffering. From this perspective, autonomous battlefield robots appear morally attractive and permissible. If their deployment leads to fewer deaths overall, then their use would be considered morally preferable to traditional combat. Supporters of this view often add that autonomous systems may outperform human soldiers in accuracy and consistency. Unlike humans, machines do not panic, hesitate, or act under emotional pressure. By applying targeting rules uniformly and without fatigue, such systems might reduce mistakes caused by human error, thereby limiting unintended harm.

Taken on its own, this outcome-focused reasoning is somewhat persuasive. If war is unavoidable, then minimising its human cost seems like a clear moral improvement. Any convincing rejection of the claim must therefore explain why this reasoning is insufficient.

One reason to doubt the claim that autonomous battlefield robots are a good thing is the risk of hacking or system compromise. Because such systems operate at speed and without direct human oversight, unauthorised access could allow hostile actors to redirect them towards allied forces or civilians. This possibility raises concerns about uncontrolled violence and undermines the assumption that autonomous systems reliably protect allied lives.

A natural response is that advances in military technology are typically accompanied by advances in cybersecurity. Encryption, redundancy, and fail-safe mechanisms could significantly reduce the likelihood of unauthorised access. From this perspective, hacking appears to be a technical challenge rather than a decisive moral objection, weakening its persuasiveness against the claim.

However, this response is a weak counter argument. No security system is entirely immune to breach, and even highly protected military and governmental infrastructure has been compromised. This restores the objection to some extent, as the possibility of hacking therefore cannot be eliminated in principle. Nevertheless, even when this is granted, hacking remains a comparatively weak reason to reject the claim. It concerns reliability rather than moral legitimacy and does not, by itself, show that preventing allied deaths is insufficient justification.

A more significant challenge to the claim arises once the utilitarian argument itself is examined. Although reducing casualties is morally relevant, outcome-based reasoning alone cannot determine whether the outcomes achieved are morally acceptable. An action may reduce suffering overall while still being ethically troubling in how it is carried out.

Supporters of autonomous battlefield robots may respond that war is already morally tragic and that reducing harm wherever possible should take priority. However, this response assumes that killing is merely a technical problem to be optimised. It

overlooks the fact that warfare involves deliberate acts of violence that demand justification at the point of action, not merely favourable results.

As a result, even if utilitarian calculations favour autonomous battlefield robots, this does not settle whether delegating decisions about killing is permissible. Therefore, it follows that harm reduction, while important, cannot by itself justify the conclusion that autonomous battlefield robots are a good thing.

The strongest reason to reject the claim concerns moral responsibility. In traditional warfare, lethal force is exercised by human agents who can be held accountable for their actions. Soldiers may be praised or blamed, punished or excused, depending on their intentions and conduct. This framework of responsibility plays a crucial role in constraining violence and maintaining moral limits within war.

Autonomous battlefield robots change this structure. When a machine independently selects and engages a target, responsibility becomes fragmented. The programmer did not choose the individual victim, the commander did not authorise the specific act, and the machine itself cannot bear moral responsibility. Who can we hold responsible? This creates a responsibility gap in which lethal harm occurs without a clearly accountable agent (Sparrow, 2016). Unlike outcome-based objections, this argument targets the moral legitimacy of lethal decision-making itself and therefore directly challenges the inference from “fewer allied deaths” to “a good thing”.

The importance of human moral agency becomes clearer when considering cases in which refusal of orders can prevent wrongful killing. During the Second Lebanon War, Israeli soldier Benaya Rein was ordered to open fire on a group believed to be enemy combatants. Rather than acting immediately, he hesitated and chose to investigate further, later discovering that the individuals were in fact fellow soldiers. His refusal to act saved their lives (Stilwell, 2022).

It may be objected that such cases are rare and outweighed by the overall reduction in casualties autonomous systems might achieve. However, this response misunderstands what is morally significant about the example. Its value lies crucially in the capacity for moral restraint in the moment that mortal decisions are made than just the amount of people saved. An autonomous system operating purely on predefined targeting criteria would have been unable to question the order or reassess the situation. This directly undermines the claim that preventing allied deaths alone is sufficient justification.

A further defence of the original claim could be that there is no morally significant difference between a human killing and a machine killing. Warfare has always involved technological mediation, from swords to firearms to long-range missiles, and responsibility ultimately lies with those who authorise violence. On this view, whether a

target is killed by a human or an autonomous system is morally irrelevant. If killing by machines is equivalent to killing by humans, then saving allied lives appears to be good.

This argument has some strength, particularly against objections that rely solely on emotional discomfort. War is not morally justified simply because killing is distressing, nor necessarily rendered worse because killing is carried out ‘efficiently’. However, this defence fails to defeat the rejection of the claim. Even if the act of killing itself is morally equivalent, delegating lethal decisions to machines reshapes how responsibility and restraint function within war. The issue is not who pulls the trigger, but how killing is organised. As such, this defence does not rescue the claim that saving allied lives alone makes autonomous battlefield robots a good thing.

It may further be argued that sufficiently advanced autonomous systems could one day replicate human-like deliberation, hesitation, or even moral reasoning, thereby preserving ethical legitimacy. If such systems appeared conscious and capable of weighing reasons in a way indistinguishable from human judgement, the distinction between human and machine decision-making might seem to collapse.

However, this line of reasoning rests on a mistake about what grounds moral responsibility. Philosophical debates about moral agency suggest that outward behaviour and functional similarity are not sufficient. A system may convincingly imitate hesitation, remorse, or deliberation without genuinely understanding reasons or recognising moral wrongdoing. This is often referred to as the problem of consciousness: we would not treat a being as morally responsible simply because it appears human-like if its behaviour is ultimately the result of programmed responses rather than genuine moral understanding.

Even if an autonomous system perfectly simulates moral deliberation, its actions remain the product of optimisation and design choices made by others. It cannot be held responsible for its decisions, nor can it meaningfully answer for moral failure. Moral responsibility requires more than the appearance of judgement, because it also requires the capacity to respond to ethical dilemmas and to be accountable for acting upon them. Autonomous systems lack these capacities regardless of how human-like they appear. As a result, the suggestion that sufficiently advanced imitation could preserve ethical legitimacy fails to undermine the responsibility-based objection.

An even stronger objection to the claim concerns moral equality. By focusing on allied soldiers, the claim treats harm to opposing soldiers as morally secondary, even though soldiers on the opposing side are just individuals compelled to fight by political authorities. When autonomous systems are introduced asymmetrically, this moral bias is intensified. One side bears minimal risk, while the other suffers almost all the bodily harm. As autonomous technology advances, this imbalance is likely to grow.

The strongest challenge arises when considering what happens if both sides rely primarily on autonomous systems. Conflict would increasingly resemble an artificial intelligence arms race, with outcomes determined by technological capacity rather than human agency or political legitimacy. Wealthier states would gain overwhelming advantages, while poorer states would become increasingly vulnerable. At the same time, removing human risk weakens one of the strongest deterrents to war. When citizens are no longer directly exposed to harm, wars become easier to start and harder to stop. Even if allied deaths are reduced, the transformation of war itself weakens the claim that autonomous battlefield robots are a good thing, as it rests on an incomplete piece of reasoning. While reducing casualties is morally relevant, it does not follow that this consideration alone is sufficient to justify delegating decisions about killing to machines. This logical gap is exposed by utilitarian appeals to harm reduction, which, although powerful, fail to account for how lethal force is exercised and who bears responsibility for it.

In conclusion, preventing allied soldiers from dying needlessly may count in favour of autonomous systems, but it is not inherently sufficient to make them a good thing. Saving lives alone is not enough to justify reshaping war in ways that remove human accountability from decisions to kill. Autonomous battlefield robots undermine moral responsibility at the point of lethal decision-making, distort the moral equality of combatants by concentrating risk on those with the least agency, and transform war into a technological arms race detached from human restraint. These considerations outweigh the harm-reduction argument, and they show why the argument cannot do the justificatory work the claim demands.

Bibliography

← All essays