Linda Eggert

This paper sketches elements of a theory of the ethics of autonomised harming: the phenomenon of delegating decisions about whether and whom to harm to artificial intelligence (AI) in self-driving cars and autonomous weapon systems. First, the paper elucidates the challenge of integrating non-human, artificial agents, which lack rights and duties, into our moral framework which relies on precisely these notions to determine the permissibility of harming. Second, the paper examines how potential differences between human agents and non-human, artificial agents might bear on the permissibility of delegating life-and death decisions to AI systems. Third, and finally, the paper explores a series of resulting complexities. These include the challenge of weighing autonomous systems’ promise to reduce harm against the intrinsic value of rectificatory justice as well as the peculiar possibility that delegating harmful acts to AI might render ordinarily impermissible acts permissible. By illuminating what happens when we extend normative theory beyond its traditional boundaries, this discussion offers a starting point for assessing the moral permissibility of delegating consequential decisions to non-human, artificial agents.

This paper defends the possibility of dirty hands against the longstanding skepticism that an action cannot be simultaneously right and wrong and that dirty hands cases are therefore impossible. While skeptics are right to recognize that prima facie reasons against violating moral duties may be overridden, they are wrong to deny that actions required by necessity may nevertheless remain wrong. Dirty hands cases capture the simultaneous necessity of disregarding moral duties in certain circumstances and the reprehensibility of wronging people even in cases in which this is all-things-considered permissible. Rather than indicating any deficiency in our moral reasoning, the capacity to accommodate the possibility of dirty hands, along with the reality of moral conflict, is a strength of a moral theory.

Recent debates about supererogatory rescues have sought to explain how it can be wrong to perform a suboptimal rescue although it would be permissible not to rescue at all. This paper proposes a new solution to this puzzle. It argues that existing accounts have neglected two critical considerations. First, contrary to what is commonly assumed, a rescue’s supererogatory nature has no bearing on the duties that apply to agents who rescue in supererogatory fashion. Second, we cannot justify harms caused as a side effect of supererogatory rescues by appealing to the fact that it would have been permissible not to rescue at all. Ultimately, the paper proposes, the same duties that apply in cases in which rescuing is required also apply in cases in which rescuing is supererogatory. A rescue’s supererogatory nature, it turns out, is not the game changer we thought it was.

This paper examines what prevents us from legally enforcing the moral imperative of protecting human rights during military operations carried out for distinctly humanitarian purposes. The answer, I argue, lies not in familiar objections to bringing the law into greater congruence with morality, but in international law's indeterminacy regarding the use of force. Preserving stability within the nascent international legal system comes at the cost of a law that eschews the protection of individual rights even in cases in which the protection of human rights is what justifies military action. The tension between state sovereignty and the protection of human rights thus not only generates well-known controversies about the lawfulness of military intervention. It also prevents us from devising laws to protect human rights during wars whose very purpose it is to stop human rights violations. Protecting human rights during humanitarian interventions may thus remain an undertaking as quixotic as it is morally urgent.

This paper defends revisionism against a challenge: that it cannot convincingly hold that many instances of killing in war are morally wrong but should nonetheless remain legally permissible. The paper argues that we should view the relationship between the morality of war and the laws of war as analogous to the relationship between fundamental principles and rules of regulation in debates about theories of justice. This yields a fresh justification for the law’s divergence from morality, which absolves revisionism from the twofold charge of being at once irrelevant and incoherent. It also helps clarify persisting ambiguities concerning the contrast between war and ‘ordinary life,’ the contested notion of a ‘deep morality’ of war, and revisionism’s relationship to ideal and non-ideal theory.

This paper examines whether the prospect of compensation may render otherwise disproportionate harms proportionate. It argues that we should reject this possibility. Instead, it distinguishes duties of compensation as a requirement of rectificatory justice from a harm’s degree of compensability, and argues that only the latter is relevant to proportionality. On this view, failing to compensate constitutes a distinct wrong, while harms that are not adequately compensable carry extra weight in proportionality calculations. This explains how the prospect of compensation affects permissions to harm, without relying on the dubious assumption that disproportionate harms may be rendered proportionate by future compensation.

This paper challenges one line of argument which has been advanced to justify imposing risks of collateral harm on prospective beneficiaries of armed humanitarian interventions. This argument - the ‘Beneficiary Principle’ - holds that non-liable individuals’ immunity to being harmed as a side effect of just armed humanitarian interventions may be diminished by their prospects of benefiting from the intervention. Against this, I defend the view that beneficiary status does not morally distinguish beneficiaries from other non-liable individuals in such a way as to permit exposing them to greater risks of being harmed. The argument proceeds in four steps. I first show that the BP can neither be grounded in liability-based nor in lesser-evil justifications for harming. I then argue that a standalone justification for unintended harming based on beneficiary status would face at least two critical challenges. The first concerns the BP’s applicability to collectives; the second questions the normative weight we can plausibly ascribe to beneficiary status when beneficiaries are such by virtue of being victims of wrongful threats of harm. I argue that standing to benefit is morally irrelevant when the benefit consists in the mitigation or prevention of wrongful harms, and consequently suggest that the BP may only serve as a distributive principle in allocating risks of harm if it is disambiguated in a number of critical aspects and applied in a more narrowlydefined set of circumstances.