This report examines the issue of human control with regard to lethal autonomy, an issue of significant interest in United Nations discussions in the Convention on Certain Conventional Weapons (CCW) forum. We analyze this issue in light of lessons and best practices from recent U.S. operations. Based on this analysis, we make the case for a wider framework for the application of human control over the use of force. This report recommends that CCW discussions currently focusing on process considerations, such as human control, should instead focus on outcome—namely, mitigation of inadvertent engagements. This allows consideration of a more complete set of benefits and risks of lethal autonomy and better management of risks. The report also describes best practices that can collectively serve as a safety net for the use of lethal autonomous weapons. It concludes with concrete recommendations for how the international community can more effectively address the risk of inadvertent engagements from lethal autonomy.
View the full report from the Center for Autonomy and AI (CAAI) here.