Nathan Gabriel Wood

Across academia, industry, and government, “AI” has become central in research and development, regulatory debates, and promises of ever faster and more capable decision-making and action. In numerous domains, especially safety-critical ones, there are significant concerns over how “AI” may affect decision-making, responsibility, or the likelihood of mistakes (to name only a few categories of critique). However, for most critiques, the target is generally “AI”, a broad term admitting many (types of) systems used for a variety of tasks and each coming with its own set of limitations, challenges, and potential use cases. In this article, we focus on the military domain as a case study and present both a loose enumerative taxonomy of systems captured under the umbrella term “military AI”, as well as discussion of the challenges of each. In doing so, we highlight that critiques of one (type of) system will not always transfer to other (types of) systems. Building on this, we argue that in order for debates to move forward fruitfully, it is imperative that the discussions be made more precise and that “AI” be excised from debates to the extent possible, especially as discussions move more toward discrete systems and use cases. Researchers, developers, and policy-makers should make clear exactly what systems they have in mind and what possible benefits and risks attend the deployment of those particular systems. While we focus on AI in the military as an exemplar for the overall trends in discussions of “AI”, the argument’s conclusions are broad and have import for discussions of AI across a host of domains.

Across academia, industry, and government, "AI" has become central in research and development, regulatory debates, and promises of ever faster and more capable decision-making and action. In numerous domains, especially safety-critical ones, there are significant concerns over how "AI" may affect decision-making, responsibility, or the likelihood of mistakes (to name only a few categories of critique). However, for most critiques, the target is generally "AI", a broad term admitting many (types of) systems used for a variety of tasks and each coming with its own set of limitations, challenges, and potential use cases. In this article, we focus on the military domain as a case study and present both a loose enumerative taxonomy of systems captured under the umbrella term "military AI", as well as discussion of the challenges of each. In doing so, we highlight that critiques of one (type of) system will not always transfer to other (types of) systems. Building on this, we argue that in order for debates to move forward fruitfully, it is imperative that the discussions be made more precise and that "AI" be excised from debates to the extent possible. Researchers, developers, and policy-makers should make clear exactly what systems they have in mind and what possible benefits and risks attend the deployment of those particular systems. While we focus on AI in the military as an exemplar for the overall trends in discussions of "AI", the argument's conclusions are broad and have import for discussions of AI across a host of domains.

Critics of autonomous weapons systems (AWS) argue that these weapons cannot reliably distinguish between legitimate targets and those protected from attack. As a result, the use of AWS seems to violate the principle of distinction under international humanitarian law (IHL), which requires that combatants “not make civilians the object of attack” and not carry out attacks that are “indiscriminate in nature.” This criticism, however, misunderstands the principle of distinction and ignores important aspects of how AWS are being developed and deployed. Critics rely on an overly broad definition of AWS, and hold these systems to a standard that is inconsistent with the principle of distinction as it is actually formulated under IHL. Despite their very real limitations, the characteristics of modern AWS in fact highlight an impressive feat of technological design, and mark a further step on our long and fitful road to making warfare a less brutal and bloody enterprise.

A weapon is a weapon, and nothing about its general “weaponness” seems to indicate much about its underlying values. However, closer examination of weapons’ designs can tell us much. For whom were they designed? Have safeguards been put into place? If so, which ones, and to the protection of whom? Autonomous weapons have existed for decades, and AI-powered and AI-enabled systems are becoming increasingly important, but all of these systems raise ethical and legal concerns. Arguably, many of the challenges might feasibly be addressed through design or institutional solutions, but importantly, how systems are designed says a great deal about those designing them and those contemplating their deployment. In this chapter, I argue that if states wish to demonstrate their commitment to human rights and the protection of innocent lives, autonomous and AI-enabled systems in the military must fulfill several design desiderata: predictability, governability, responsibility, and traceability. I then explore concrete ways these desiderata may be fulfilled in practice.

Use of autonomous, AI-enabled, or opaque systems raises many concerns, and some argue that for these to be permissibly deployed in high-stakes or critical domains, they must be subject to so-called “meaningful human control” (MHC). In this article, I focus on the military domain and rebut a strong version of this critique, arguing that off-the-loop systems – i.e., those which can select and engage targets without contemporaneous human input or oversight – can be permissibly deployed while retaining clear lines of responsibility and control. I show that ex ante operational constraints and targeting parameters can provide combatants and would-be deployers of off-the-loop systems with strong means to ensure that deployed systems are serving as extensions of combatants’ wills, establishing the necessary degree of moral and legal responsibility required. I further show that such constraints and parameters represent clear lines of control that deployers have over such systems, even when these systems, during deployments, are utterly outside of human control. I conclude by distinguishing between what I call “will-extending” and “will-offloading” systems, showing that off-the-loop systems can serve to extend users’ and deployers’ wills, making such systems inherently subject to meaningful human control. Throughout, the discussion focuses on the example of autonomous and AI-enabled systems in the military domain, but the underlying arguments relate to such systems more generally, showing how these, if utilized as “will-extending” systems, may be used in a controlled and responsibility-retaining manner.

Use of autonomous, AI-enabled, or opaque systems raises many concerns, and some argue that for these to be permissibly deployed in high-stakes or critical domains, they must be subject to so-called “meaningful human control” (MHC). In this article, I focus on the military domain and rebut a strong version of this critique, arguing that off-the-loop systems – i.e., those which can select and engage targets without contemporaneous human input or oversight – can be permissibly deployed while retaining clear lines of responsibility and control. I show that ex ante operational constraints and targeting parameters can provide combatants and would-be deployers of off-the-loop systems with strong means to ensure that deployed systems are serving as extensions of combatants’ wills, establishing the necessary degree of moral and legal responsibility required. I further show that such constraints and parameters represent clear lines of control that deployers have over such systems, even when these systems, during deployments, are utterly outside of human control. I conclude by distinguishing between what I call “will-extending” and “will-offloading” systems, showing that off-the-loop systems can serve to extend users’ and deployers’ wills, making such systems inherently subject to meaningful human control. Throughout, the discussion focuses on the example of autonomous and AI-enabled systems in the military domain, but the underlying arguments relate to such systems more generally, showing how these, if utilized as “will-extending” systems, may be used in a controlled and responsibility-retaining manner.

On a contentious but still widely held view of the ethics of war, belligerents’ mutual imposition of risk licenses the harm they attempt to inflict upon each other. When this reciprocity of risk imposition is lost—when combatants of one side are able to inflict harm without exposing themselves to it—the moral balance is disrupted. Technologies that radically reduce risk, such as UAVs (drones) or autonomous weapon systems, are particularly challenging in this respect. Scholars have suggested that these technologies of radical risk reduction are so morally disruptive that the ethics of war should fundamentally shift from virtue-based towards rule-based approaches. According to this view, since the traditional martial virtues (such as courage and mercy) are of little practical relevance to certain forms of modern warfare, drone pilots and other soldiers who dole out violence from great distances would be better served by clear moral rules. We argue, however, that this view is mistaken. That technologies of radical risk reduction bring some disruption to the martial virtues is undeniable, but the disruption is far less severe than has been suggested. We present an alternative account of the moral disruption caused by technologies of radical risk-reduction, highlighting how the changing nature of warfare is informing changes in what morality and virtue demand in war. Martial virtues continue to play a crucial role in ethical decision-making, as the moral performance of drone pilots in Ukraine has clearly demonstrated. We continue to have, therefore, strong reasons to educate and entrench a conception of the ‘virtuous warrior’ in today’s soldiers.

Large language models (LLMs) such as ChatGPT present immense opportunities, but without proper training for users (and potentially oversight), they carry risks of misuse as well. We argue that current approaches focusing predominantly on transparency and explainability fall short in addressing the diverse needs and concerns of various user groups. We highlight the limitations of existing methodologies and propose a framework anchored on user-centric guidelines. In particular, we argue that LLM users should be given guidelines on what tasks LLMs can do well and which they cannot, which tasks require further guidance or refinement by the user, and context-specific heuristics. We further argue that (some) users should be taught to refine and elaborate adequate prompts, be provided with good procedures for prompt iteration, and be taught efficient ways to verify outputs. We suggest that for users, shifting away from looking at the technology itself, but rather looking at the usage of it within contextualized sociotechnical systems, can help solve many issues related to LLMs. We further emphasize the role of real-world case studies in shaping these guidelines, ensuring they are grounded in practical, applicable strategies. Like any technology, risks of misuse can be managed through education, regulation, and responsible development.

Artificial intelligence (AI) has become nearly ubiquitous in modern society, from components of mobile applications to medical support systems, and everything in between. In societally impactful systems imbued with AI, there has been increasing concern related to opaque AI, that is, artificial intelligence where it is unclear how or why certain decisions are reached. This has led to a recent boom in research on “explainable AI” (XAI), or approaches to making AI more explainable and understandable to human users. In the military domain, numerous bodies have argued that autonomous and AI-enabled weapon systems ought not incorporate unexplainable AI, with the International Committee of the Red Cross and the United States Department of Defense both explicitly including explainability as a relevant factor in the development and use of such systems. In this article, I present a cautiously critical assessment of this view, arguing that explainability will be irrelevant for many current and near-future autonomous systems in the military (which do not incorporate any AI), that it will be trivially incorporated into most military systems which do possess AI (as these generally possess simpler AI systems), and that for those systems with genuinely opaque AI, explainability will prove to be of more limited value than one might imagine. In particular, I argue that explainability, while indeed a virtue in design, is a virtue aimed primarily at designers and troubleshooters of AI-enabled systems, but is far less relevant for users and handlers actually deploying these systems. I further argue that human–machine teaming is a far more important element of responsibly using AI for military purposes, adding that explainability may undermine efforts to improve human–machine teamings by creating aprima faciesense that the AI, due to its explainability, may be utilized with little (or less) potential for mistakes. I conclude by clarifying that the arguments are not against XAI in the military, but are instead intended as a caution against over-inflating the value of XAI in this domain, or ignoring the limitations and potential pitfalls of this approach.

Worsening energy crises and the growing effects of climate change have spurred, among other things, concerted efforts to tackle global problems through what the United Nations calls Sustainable Development Goals (SDGs). These are in turn argued to be best achieved via the adoption of environmental, social, and corporate governance (ESG) as the vehicle for guiding our efforts. However, though these things are often presented as the solution to global issues, they are increasingly being used as a means to centralize power and exacerbate growing authoritarian trends. This article aims to demonstrate how the narratives surrounding ESG mask this reality, taking as a solution one of the foundational causes of global wealth and power disparity.

The principle of proportionality demands that a war (or action in war) achieve more goods than bads. In the philosophical literature there has been a wealth of work examining precisely which goods and bads may count toward this evaluation. However, in all of these discussions there is no mention of one of the most certain bads of war, namely the psychological harm(s) likely to be suffered by the combatants who ultimately must fight and kill for the purposes of winning in conflict. This paper argues that harms to one’s own soldiers must be included in proportionality judgments, and goes on to argue that one of the most significant harms one’s soldiers face are the psychological stresses and traumas associated with combat. The arguments draw on a growing wealth of psychological literature exploring the connections between combatancy and psychological trauma, and highlight, in particular, the uniquely negative impact which killing has on a combatant’s mental well-being. The paper concludes that these factors place an almost certain and rather weighty negative weight in any proportionality calculations concerning wars with ground combatants who must fight “up close and personal”, and that for more remote warfighters, there is also evidence to show that they may suffer deep psychological harm as a result of their combat roles as well. The argument, however, does not attempt to demonstrate that these factors render war impermissible. Rather, it merely shows that these harms, or bads, which can be quite significant, must be factored into our considerations of proportionality. The arguments themselves are rather uncontroversial, but they bring to light an element in the moral calculus which is sadly overlooked in most discussions of the ethics of war.

In his article, “Should Space Travel be Human or Robotic? Reasons for and against full automation for space missions,” Maurizio Balistreri explores the ongoing debate regarding whether space travel, exploration, and extra-terrestrial colonization should be the domain of humans or robots. Balistreri explores both technical and normative arguments for why extraterrestrial ventures ought to be wholly robotic or human, ultimately taking no explicit side in the debate. However, in this article we argue that by even posing the question in this binary fashion, Balistreri and others are making a mistake at the outset. We show that there are certain missions and aims which require humans and others that plausibly may be undertaken solely by robots, but for all missions and aims there is likely to be some degree of human-robot pairing. More than this, we show that the question should not be whether space travel, exploration, or extra-terrestrial colonization ought to be generally human-centric or robot-centric, but rather that each and every mission should be examined on its own, as the values and disadvantages of humans versus robots are apt to be highly specific to the realities of particular discrete missions.

Due to advances in military technology, there has been an outpouring of research on what are known as autonomous weapon systems (AWS). However, it is common in this literature for arguments to be made without first making clear exactly what definitions one is employing, with the detrimental effect that authors may speak past one another or even miss the targets of their arguments. In this article I examine the U.S. Department of Defense and International Committee of the Red Cross definitions of AWS, showing that these definitions are far broader than some recognize, and that they therefore classify a much larger set of weapons as AWS. I then show that these broader views of AWS have implications for what moral and legal rules we might argue should be applied to such systems. I conclude by arguing there is a greater need for precision and clarity within AWS debates, in order to ensure that researchers are discussing the same weapon systems (autonomous or otherwise) when they argue for or against particular points. The purpose of this article is not to defend any specific view of AWS, nor to further any general endorsement or objection to such systems, but rather to show the importance of argumentative clarity in this debate.

A classic objection to autonomous weapon systems (AWS) is that these could create so-called responsibility gaps, where it is unclear who should be held responsible in the event that an AWS were to violate some portion of the law of armed conflict (LOAC). However, those who raise this objection generally do so presenting it as a problem for AWS as a whole class of weapons. Yet there exists a rather wide range of systems that can be counted as “autonomous weapon systems”, and so the objection is too broad. In this article I present a taxonomic approach to the objection, examining a number of systems that would count as AWS under the prevalent definitions provided by the United States Department of Defense and the International Committee of the Red Cross, and I show that for virtually all such systems there is a clear locus of responsibility which presents itself as soon as one focuses on specific systems, rather than general notions of AWS. In developing these points, I also suggest a method for dealing with near-future types of AWS which may be thought to create situations where responsibility gaps can still arise. The main purpose of the arguments is, however, not to show that responsibility gaps do not exist or can be closed where they do exist. Rather, it is to highlight that any arguments surrounding AWS must be made with reference to specific weapon platforms imbued with specific abilities, subject to specific limitations, and deployed to specific times and places for specific purposes. More succinctly, the arguments show that we cannot and should not aim to treat AWS as if all of these shared all morally relevant features, but instead on a case-by-case basis. Thus, we must contend with the realities of weapons development and deployment, and tailor our arguments and conclusions to those realities, and with an eye to what facts obtain for particular systems fulfilling particular combat roles.

In international law and the ethics of war, there are a variety of actions which are seen as particularly problematic and presumed to be always or inherently wrong, or in need of some overwhelmingly strong justification to override the presumption against them. One of these actions is assassination, in particular, assassination of heads of state. In this essay I argue that the presumption against assassination is incorrect. In particular, I argue that if in a given scenario war is justified, then assassination of the enemy state's leader is also justified, and in fact ought to be pursued as a means short of war. I defend this position on both consequential and deontic grounds, arguing that assassination is both more discriminate than war and serves to harm only those most responsible for the situation which justifies war in the first place. I conclude by arguing that a norm of assassination, far from being a destabilizing force, as some have argued, would in fact serve to reinforce international norms of rights and respect for persons by making clear that tyrants and would-be oppressors cannot hide behind military forces or notions of sovereignty to protect themselves from judgment and retribution.

Warfare is becoming increasingly automated, from automatic missile defense systems to micro-UAVs (WASPs) that can maneuver through urban environments with ease, and each advance brings with it ethical questions in need of resolving. Proponents of lethal autonomous weapons systems (LAWS) provide varied arguments in their favor; robots are capable of better identifying combatants and civilians, thus reducing "collateral damage"; robots need not protect themselves and so can incur more risks to protect innocents or gather more information before using deadly force; robots can assess situations more quickly and do so without emotion, reducing the likelihood of fatal mistakes due to human error; and sending robots to war protects our own soldiers from harm. However, these arguments only point in favor of autonomous weapons systems, failing to demonstrate why such systems need be made *lethal*. In this paper I argue that if one grants all of the proponents' points in favor of LAWS, then, contrary to what might be expected, this leads to the conclusion that it would be both immoral and illegal to deploy *lethal* autonomous weapons, because the many features that speak in favor of them also undermine the need for them to be programmed to take lives. In particular, I argue that such systems, if lethal, would violate the moral and legal principle of necessity, which forbids the use of weapons that impose superfluous injury or unnecessary harm. I conclude by highlighting that the argument is not against autonomous weapons per se, but only against *lethal* autonomous weapons.

Autonomous weapons systems (AWS), sometimes referred to as “killer robots”, are receiving evermore attention, both in public discourse as well as by scholars and policymakers. Much of this interest is connected with emerging ethical and legal problems linked to increasing autonomy in weapons systems, but there is a general underappreciation for the ways in which existing law might impact on these new technologies. In this paper, we argue that as AWS become more sophisticated and increasingly more capable than flesh-and-blood soldiers, it will increasingly be the case that such soldiers are “in the power” of those AWS which fight against them. This implies that such soldiers ought to be considered hors de combat, and not targeted. In arguing for this point, we draw out a broader conclusion regarding hors de combat status, namely that it must be viewed contextually, with close reference to the capabilities of combatants on both sides of any discreet engagement. Given this point, and the fact that AWS may come in many shapes and sizes, and can be made for many different missions, we argue that each particular AWS will likely need its own standard for when enemy soldiers are deemed hors de combat. We conclude by examining how these nuanced views of hors de combat status might impact on meaningful human control of AWS.

Contemporary proponents of republican political theory often focus on the concept of freedom as non-domination, and how best to promote it within a state. However, there is little attention paid to what the republican conception of freedom demands in the international realm. In this essay I examine what is required for an agent to enjoy freedom as non-domination, and argue that this might only be achieved for individuals if one of two possibilities is pursued internationally: either (1) all nations are made equally powerful, such that none may arbitrarily impose its will on another without penalty, or (2) all nations are joined under a global sovereign which guarantees that the weaker states are not subject to the whims of the stronger. I further argue that the first condition cannot suffice for achieving true non-domination, and as a result, republicanism must prescribe working toward the establishment of a global state. This paper provides an important contribution to the literature by addressing the international implications of the republican conception of freedom.

In a recent article Duncan Purves, Ryan Jenkins, and B. J. Strawser argue that in order for a decision in war to be just, or indeed the decision to resort to war to be just, it must be the case that the decision is made for the right reasons. Furthermore, they argue that this requirement holds regardless of how much good is produced by said action. In this essay I argue that their argument is flawed, in that it mistakes what makes an agent morally good for what makes an act morally good. I argue that the main thrust of Purves et al.’s argument in fact undermines the conclusion they wish to draw, and that the reasons for one’s action do not make an in principle difference to the morality of actions in war. I further argue that this position undermines the traditional ad bellum just war constraint of right intention, and that the morality of actions in war is, at core, only concerned with outcomes. I conclude by clarifying that one’s reasons for action do in fact matter when deciding to enter war or kill in war, but only because one’s reasons significantly impact the way in which one acts. The purpose of this paper is to clear the theoretical space by showing why intentions/reasons do not in principle matter when assessing the morality of war (or killing), but this should not be taken as an argument that we should ignore intentions/reasons altogether.