Beyond Automation: The Ethical, Doctrinal, and Operational Challenges of Human-AI Collaboration in Military Decision-Making

Abstract

The accelerating integration of artificial intelligence (AI) into military systems presents unprecedented opportunities and complex dilemmas, particularly in the domain of decision-making under pressure. This article critically explores the ethical, doctrinal, and operational challenges that arise when human-AI collaboration is employed in high-stakes military contexts. At the heart of the study is the question of how military organizations can construct human-AI teams that simultaneously enhance operational effectiveness, uphold legal and ethical standards, and maintain meaningful human control over critical functions. To address this challenge, the study adopts a qualitative and interdisciplinary methodology that synthesizes doctrinal analysis, particularly of NATO and U.S. Department of Defense frameworks, with in-depth case studies of AI-enabled systems used in military operations. It investigates how AI systems affect traditional decision-making processes by accelerating data synthesis, improving situational awareness, and enabling faster reaction times. However, it also highlights emerging risks, such as automation bias, the erosion of human moral agency, loss of interpretability of algorithmic decisions, and the increasing difficulty in assigning responsibility for outcomes. The findings underscore that without appropriate safeguards, AI could undermine ethical accountability and legal clarity. To mitigate these risks, the article proposes a comprehensive framework for designing effective human-AI teams, which includes the implementation of transparent system architectures, explainable AI (XAI) models, trust calibration strategies, adaptive training modules, and multi-layered oversight mechanisms. Special attention is given to the necessity of doctrinal adaptation, the cultural and institutional readiness of military organizations, and the role of normative principles in regulating machine autonomy. Ultimately, the article argues that responsible innovation in military AI must be grounded in ethical rigor, legal certainty, and strategic prudence. Only by embedding these values into the design, deployment, and governance of human-AI teaming can military institutions ensure that AI enhances, rather than compromises, the legitimacy and effectiveness of defense operations.