Trustworthy Human-Autonomy Teaming for Proportionality Assessment in Military Operations

Over the past decades, rapid technological advancements resulted in the integration of autonomous systems and AI across various societal domains. An emerging paradigm in this realm is the human-autonomy teaming which merges human efforts and intelligence with efficiency and performance of autonomous systems through collaboration for reaching common goals and leveraging their strengths. Building such systems should be done in a safe, responsible, and reliable manner, thus in a trustworthy way. While efforts for developing such systems exist in the military domain, they are mainly tackling the technical dimension involved in tasks like reconnaissance and target engagement, and less in conjunction with other dimensions like ethical and legal while focusing on possible produced effects. The intended effects contributing to achieving military objectives represent military advantage and the unintended effects on civilian and civilian objects represent collateral damage. Bringing and assessing these types of effects in a single instance is done through the proportionality assessment which represents the pilar when conducting military operations. Nevertheless, no previous efforts are dedicated to building human-autonomy teaming systems in the context of proportionality assessment in military operations in a trustworthy way. Hence, it is the aim of this research to define this concept and propose a corresponding design framework on this behalf with the intention to contribute to building safe, responsible, and reliable systems in the military domain. To achieve this goal, the Design Science Research methodology is followed in a Value Sensitive Design approach based on extensive research of relevant studies and field experience.