An Air Force officer at USU studies how AI decision tools in combat could shape moral injury and the care service members will need.
A USU graduate student with a background in moral philosophy is preparing to study moral injury in an era of AI-enabled warfare. U.S. Air Force 2nd Lt. Raymond Gilmore, a graduate student in the Department of Medical and Clinical Psychology (MPS) at the Uniformed Services University of the Health Sciences (USU), is undertaking a crucial study into the intersection of artificial intelligence (AI) and moral injury in modern warfare. Prior to his military psychology training, Gilmore immersed himself in moral philosophy and and theology at the Catholic University of America, providing him with a robust ethical framework for his current research. His work aims to understand how the increasing integration of AI into military command-and-control operations will alter the psychological burdens, specifically moral injuries, that service members might experience upon returning from combat. This initiative is a core part of USU's broader strategic focus on the ethical implications and practical applications of AI in military medicine. Gilmore's philosophical foundation allows him to tackle fundamental questions about human nature and consciousness within the context of war, asserting that a deep understanding of these concepts is essential before effectively discussing AI's role. His research specifically addresses the lasting psychological harm, characterized by profound guilt and shame, that arises when an individual's core moral beliefs are violated by their actions or inactions in a high-stakes environment like the battlefield. By exploring these complex issues, USU is actively preparing its future healthcare providers to effectively support service members grappling with the novel ethical dilemmas posed by AI-enabled combat. Calibrating Trust: Automation Bias vs. Algorithm Aversion This section highlights a critical area of research within USU’s forward-looking approach: understanding and preparing future military healthcare providers for the complex psychological dynamics between service members and Decision Support Systems (DSS). DSS refers to sophisticated algorithmic software designed to process vast amounts of battlefield data and offer operational recommendations, including those pertaining to lethal engagement. Through high-fidelity simulations involving active-duty military personnel, researchers have observed two distinct and often contradictory cognitive responses to these AI-driven systems. The first is "automation bias," where operators tend to over-rely on the machine's recommendations, perceiving them as cognitive shortcuts, and potentially abdicating their own critical judgment. The second is "algorithm aversion," characterized by a skepticism or outright rejection of algorithmic suggestions, even when the system demonstrates reasonable accuracy or a proven track record, simply due to a preference for human decision-making over machine output. Gilmore’s analysis of these simulations reveals that trust in DSS is not static but dynamically influenced by operational factors. For instance, operators exhibited heightened algorithm aversion when the perceived risk of collateral damage to civilians was high or when they experienced fatigue. Conversely, when the system demonstrated high perceived competence, it often triggered automation bias. A significant challenge identified is the "incomplete AI explainability." While operators showed greater trust when the system provided even minimal reasoning for its recommendations, complex machine-learning models often struggle to articulate their decision-making processes transparently. This lack of transparency can leave service members with profound emotional and ethical complexities after executing machine-assisted choices, contributing to moral injury. Addressing Moral Injury and "Distantiated Intimacy" This final section delves into the profound implications of AI for moral responsibility and the unique forms of psychological distress it may engender, particularly moral injury. Gilmore emphasizes that while current military doctrine generally maintains a "human in the loop" for final decisions, the rapid global evolution of autonomous systems necessitates a re-evaluation of operational standards. He underscores the fundamental difference in the experience of responsibility between a service member who personally assesses a situation and makes a decision, and one who authorizes a course of action recommended by an AI system. This diffusion of responsibility onto a nonhuman entity can result in a distinct "flavor" of guilt and shame, profoundly shaping the moral injury experienced. The article also challenges the common misconception that remote, technology-driven warfare leads to detachment or emotional coldness. Gilmore introduces the concept of "distantiated intimacy," explaining that despite physical distance from the battlefield, operators of these systems often develop a deep emotional connection due to prolonged, close observation of events. This intimate yet remote engagement ensures that the psychological toll remains "1,000 percent still there," manifesting in unique symptom clusters centered around guilt and shame. For Gilmore and his peers in the MPS program, addressing these future challenges in military medicine requires proactive discussions, continuous study, and the development of adaptable care strategies. The goal is to ensure that healthcare providers remain flexible and competent in an ever-evolving technological landscape, ready to support warfighters through the complex ethical terrain of AI-enabled combat.
A USU graduate student with a background in moral philosophy is preparing to study moral injury in an era of AI-enabled warfare.
U.S. Air Force 2nd Lt. Raymond Gilmore, a graduate student in the Department of Medical and Clinical Psychology (MPS) at the Uniformed Services University of the Health Sciences (USU), is undertaking a crucial study into the intersection of artificial intelligence (AI) and moral injury in modern warfare. Prior to his military psychology training, Gilmore immersed himself in moral philosophy and and theology at the Catholic University of America, providing him with a robust ethical framework for his current research. His work aims to understand how the increasing integration of AI into military command-and-control operations will alter the psychological burdens, specifically moral injuries, that service members might experience upon returning from combat. This initiative is a core part of USU's broader strategic focus on the ethical implications and practical applications of AI in military medicine. Gilmore's philosophical foundation allows him to tackle fundamental questions about human nature and consciousness within the context of war, asserting that a deep understanding of these concepts is essential before effectively discussing AI's role. His research specifically addresses the lasting psychological harm, characterized by profound guilt and shame, that arises when an individual's core moral beliefs are violated by their actions or inactions in a high-stakes environment like the battlefield. By exploring these complex issues, USU is actively preparing its future healthcare providers to effectively support service members grappling with the novel ethical dilemmas posed by AI-enabled combat.
Calibrating Trust: Automation Bias vs. Algorithm Aversion
This section highlights a critical area of research within USU’s forward-looking approach: understanding and preparing future military healthcare providers for the complex psychological dynamics between service members and Decision Support Systems (DSS). DSS refers to sophisticated algorithmic software designed to process vast amounts of battlefield data and offer operational recommendations, including those pertaining to lethal engagement. Through high-fidelity simulations involving active-duty military personnel, researchers have observed two distinct and often contradictory cognitive responses to these AI-driven systems. The first is "automation bias," where operators tend to over-rely on the machine's recommendations, perceiving them as cognitive shortcuts, and potentially abdicating their own critical judgment. The second is "algorithm aversion," characterized by a skepticism or outright rejection of algorithmic suggestions, even when the system demonstrates reasonable accuracy or a proven track record, simply due to a preference for human decision-making over machine output. Gilmore’s analysis of these simulations reveals that trust in DSS is not static but dynamically influenced by operational factors. For instance, operators exhibited heightened algorithm aversion when the perceived risk of collateral damage to civilians was high or when they experienced fatigue. Conversely, when the system demonstrated high perceived competence, it often triggered automation bias. A significant challenge identified is the "incomplete AI explainability." While operators showed greater trust when the system provided even minimal reasoning for its recommendations, complex machine-learning models often struggle to articulate their decision-making processes transparently. This lack of transparency can leave service members with profound emotional and ethical complexities after executing machine-assisted choices, contributing to moral injury.
Addressing Moral Injury and "Distantiated Intimacy"
This final section delves into the profound implications of AI for moral responsibility and the unique forms of psychological distress it may engender, particularly moral injury. Gilmore emphasizes that while current military doctrine generally maintains a "human in the loop" for final decisions, the rapid global evolution of autonomous systems necessitates a re-evaluation of operational standards. He underscores the fundamental difference in the experience of responsibility between a service member who personally assesses a situation and makes a decision, and one who authorizes a course of action recommended by an AI system. This diffusion of responsibility onto a nonhuman entity can result in a distinct "flavor" of guilt and shame, profoundly shaping the moral injury experienced. The article also challenges the common misconception that remote, technology-driven warfare leads to detachment or emotional coldness. Gilmore introduces the concept of "distantiated intimacy," explaining that despite physical distance from the battlefield, operators of these systems often develop a deep emotional connection due to prolonged, close observation of events. This intimate yet remote engagement ensures that the psychological toll remains "1,000 percent still there," manifesting in unique symptom clusters centered around guilt and shame. For Gilmore and his peers in the MPS program, addressing these future challenges in military medicine requires proactive discussions, continuous study, and the development of adaptable care strategies. The goal is to ensure that healthcare providers remain flexible and competent in an ever-evolving technological landscape, ready to support warfighters through the complex ethical terrain of AI-enabled combat.
