The Responsibility Gap and LAWS: a Critical Mapping of the Debate

Abstract

AI has numerous applications and in various fields, including the military domain. The increase in the degree of autonomy in some decision-making systems leads to discussions on the possible future use of lethal autonomous weapons systems (LAWS). A central issue in these discussions is the assignment of moral responsibility for some AI-based outcomes. Several authors claim that the high autonomous capability of such systems leads to a so-called “responsibility gap.” In recent years, there has been a surge in philosophical literature around the concept of responsibility gaps and different solutions have been devised to close or bridge these gaps. In order to move forward in the research around LAWS and the problem of responsibility, it is important to increase our understanding of the different perspectives and discussions in this debate. This paper attempts to do so by disentangling the various arguments and providing a critical overview. After giving a brief outline of the state of the technology of LAWS, I will review the debates over responsibility gaps using three differentiators: those who believe in the existence of responsibility gaps versus those who do not, those who hold that responsibility gaps constitute a new moral problem versus those who argue they do not, and those who claim that solutions can be successful as opposed to those who believe that it is an unsolvable problem.

Responsibility before freedom: closing the responsibility gaps for autonomous machines

“…(ii) Solutionists, by contrast, are those authors who believe that the responsibility gap can be bridged. While solutionists are united in their opposition to the pessimists, they can be differentiated in terms of the strategies they employ to close the RG, which come in a variety of flavours [30]. We will briefly discuss four:…”

“…5 An example of this solution can be integrating moral values into the design process of AMs. 6 The problem with this strategy, as pointed out by Oimann [30], is that it views the attribution of responsibility as a problem of causality, and thus fails to address the normative dimension of the RG that concerns the inability of identifying individuals that are responsible for outcomes. For example, Hidriks and Velunwenkamp argue that a certain amount of risk is acceptable in society (because it can never be reduced to zero), and if the risk associated with the use of AMs is below that tolerable threshold, then we can write off individual cases of RG as accidents for which no responsibility can be assigned.…”

“…7 To be more precise, Champagne and Tonkens' [6] proposal is that: "A person (or persons) of sufficiently high military or political standing could accept responsibility for the actions (normal or abnormal) of all autonomous robotic devices-even if that person could not be causally linked to those actions besides this prior agreement" (p. 126). 8 See p. 14 in Oimann [30] for this point in more detail. See also Kiener [18] for a more elaborate criticism of Champagne and Tonken's that centres on the (lack of) normative power of liability.…”

Conceptual Engineering and Philosophy of Technology: Amelioration or Adaptation?

“…Conceptual disruptions and adaptations need not be technology-driven: warfare, fake news, or environmental hazards might similarly give rise to social and conceptual disruptions, which call for adaptation (Oimann, 2023). Note, however, that technologies do play a significant role in each of these contexts, and the same holds for conceptual disruption more generally: concepts and conceptual schemes frequently have to be revised and adapted in the face of technological pressures.…”

“…Yet agency may not suffice for responsibility: arguably, some human-like capacity for reason responsiveness, not present in current generation AI systems, is additionally required for holding an agent responsible (see van de Poel & Sand, 2018 on conditions for ascribing responsibility). If agency is ascribed to AI systems, whereas such systems lack the requisite kind of reason responsiveness, this may result in a responsibility gap: a situation where no one can be held responsible for the harms of an autonomous system (Matthias, 2004;Oimann, 2023). Such a situation seems undesirable: we want to possess a conceptual and normative system that allows for justified ascriptions of responsibility, especially where powerful Socially Disruptive Technologies (SDTs) are at play.…”