Bad, mad, and cooked: Moral responsibility for civilian harms in human-AI military teams

TL;DR

This paper examines moral responsibility for civilian harm in military operations involving human-AI teams, proposing new frameworks to clarify responsibility and promote responsible AI deployment.

Contribution

It introduces novel mechanisms for mapping moral responsibility in human-AI military teams, including decision prompts and safety frameworks.

Findings

New decision responsibility prompts for critical decision analysis

Application of AI workplace safety frameworks to moral responsibility

Mechanisms to enable human-centered responsible AI deployment

Abstract

This chapter explores moral responsibility for civilian harms by human-artificial intelligence (AI) teams. Although militaries may have some bad apples responsible for war crimes and some mad apples unable to be responsible for their actions during a conflict, increasingly militaries may 'cook' their good apples by putting them in untenable decision-making environments through the processes of replacing human decision-making with AI determinations in war making. Responsibility for civilian harm in human-AI military teams may be contested, risking operators becoming detached, being extreme moral witnesses, becoming moral crumple zones or suffering moral injury from being part of larger human-AI systems authorised by the state. Acknowledging military ethics, human factors and AI work to date as well as critical case studies, this chapter offers new mechanisms to map out conditions for moral responsibility in human-AI teams. These include: 1) new decision responsibility prompts for critical decision method in a cognitive task analysis, and 2) applying an AI workplace health and safety framework for identifying cognitive and psychological risks relevant to attributions of moral responsibility in targeting decisions. Mechanisms such as these enable militaries to design human-centred AI systems for responsible deployment.