PMSCs meet AI-driven swarm drones: the paradigm-shifting issues of accountability

This article examines how state reliance on private military and security companies (PMSCs) affects the attribution of conduct by artificial intelligence (AI)-driven swarm drones under international law. While the outsourcing of military functions to PMSCs has long raised accountability concerns, the deployment of AI-enabled swarm technologies intensifies these challenges by introducing autonomous and adaptive decision-making that strains existing attribution frameworks. Drawing on interdisciplinary literature on AI, machine learning, and swarm intelligence, the article outlines the technological characteristics of AI-driven swarm drones and explains how their operational unpredictability contributes to a responsibility gap in international law. The analysis demonstrates that although PMSCs may complicate attribution between individuals and states, the more profound difficulty lies in attributing the conduct of autonomous systems to human actors at all. As a result, individual responsibility frameworks offer limited solutions where harm results from AI-driven swarm behaviour. The article argues that state responsibility, anchored in the obligation to conduct legal weapons reviews under Article 36 of Additional Protocol I, provides a more viable, though incomplete, mechanism for addressing accountability. It further suggests that the use of AI-enabled swarm drones by PMSCs may lower the threshold for the use of force and obscure the initiation of hostilities. The article concludes that addressing the challenges posed by AI-driven swarm drones requires a shift in emphasis towards preventive state obligations and standardized legal review practices, alongside a reassessment of how international law conceptualizes responsibility in relation to autonomous weapons systems.