Lethal autonomous weapon systems (LAWS): meaningful human Control, collective moral responsibility and institutional design

This paper investigates the ethical challenges posed by Lethal Autonomous Weapon Systems (LAWS) in military combat. The author addresses three primary concerns: the unpredictability of AI-driven systems in relation to Meaningful Human Control (MHC), the so-called 'responsibility gap' regarding who is accountable for autonomous actions, and the moral status of different levels of human-machine interaction (in-the-loop, on-the-loop, and out-of-the-loop).

To resolve these issues, the author employs a philosophical framework centered on collective moral responsibility and institutional design. By distinguishing between machine control and human moral agency, the paper argues that the unpredictability of self-learning AI does not inherently preclude moral use, provided that human controllers remain embedded in a structured institutional system that enforces accountability.

The author contends that the 'responsibility gap' is not an insurmountable barrier to the use of LAWS. Instead, it can be bridged by assigning collective institutional responsibility to the human actors—commanders, analysts, and operators—who design, deploy, and oversee these systems. Because AI lacks consciousness and the ability to act for moral reasons, it cannot be a moral agent. Therefore, moral responsibility must remain with the human beings who jointly perform the actions that lead to lethal outcomes.

Furthermore, the paper distinguishes between three levels of human-machine interaction. It concludes that 'human-out-of-the-loop' weapons should be prohibited due to their incompatibility with MHC. However, 'human-in-the-loop' and 'human-on-the-loop' systems may be morally acceptable under restricted conditions, such as when the weapon's use is governed by clear Rules of Engagement (ROE) and when human controllers retain the practical ability to override the system's lethal decisions.

As militaries increasingly integrate AI into combat, the question of whether machines can replace human judgment becomes critical. This paper provides a normative roadmap for policymakers and military leaders, suggesting that the ethical deployment of LAWS depends less on the perfection of the AI itself and more on the robustness of the institutional structures that govern its use. It emphasizes that technology must be designed to support, rather than replace, the moral discernment of human commanders.