Decision-Making Under Uncertainty in AI-Enabled Warfare: Implications for Education and Training

In AI-enabled warfare, military medical officers may be required to make life-and-death decisions when artificial intelligence (AI) outputs are correct, incorrect, or uncertain, often without the time or ability to fully verify them. As modern operations increasingly emphasize decision dominance, AI is becoming more integrated into operational environments. Although its role in command and control has been widely examined, its implications for military medical decision-making remain less defined, creating a potential gap in preparedness. Military medical decisions such as triage, evacuation, and resource allocation are inherently time-sensitive and high-stakes, often occurring in austere environments with incomplete information and competing mission priorities. In these contexts, AI may expand decision options, improve efficiency, and reduce some cognitive burdens, while also introducing complexity related to interpreting outputs, integrating multiple inputs, and calibrating trust in real time. This commentary adopts a dialectical perspective, arguing that AI may both support and complicate decision-making. Rather than eliminating uncertainty, AI redistributes it across three domains: clinical-operational, algorithmic, and relational. This redistribution introduces risks, including miscalibrated reliance, cognitive strain from competing inputs, limited transparency, and ambiguity in accountability. These dynamics raise important considerations for military medical education. Foundational training should prepare clinicians to interpret AI-supported information, integrate data sources, and maintain judgment under uncertainty. Ultimately, readiness will depend not only on individual expertise, but also on alignment among technology, training, doctrine, and system design.