Executive Summary
This study investigates what shapes US military officers' trust in AI for strategic-level decision-making. Using a conjoint experiment with senior officers, it finds that trust is influenced by technical specifications (non-lethal use, high precision, human oversight), perceived effectiveness (minimizing casualties, maximizing mission success), and international regulatory oversight. The findings challenge assumptions of automatic military adoption of AI, highlighting the complexity of trust.
Why It Matters
This document provides critical empirical evidence on how senior US military officers perceive and trust AI in strategic decision-making, directly impacting future AI integration and doctrine development. Defense analysts should care as it reveals the nuanced human element in AI adoption within military contexts, influencing policy and modernization efforts.
Key Takeaways
- Military trust in AI for strategic decisions is not automatic but shaped by specific technical, operational, and oversight factors.
- Officers prioritize AI for non-lethal purposes, with maximum precision, and with significant human oversight, distrusting fully autonomous systems.
- Trust is significantly influenced by AI's perceived effectiveness in minimizing civilian casualties and maximizing mission success, as well as international regulatory oversight.
- Congressional approval for AI use also enhances military trust, suggesting political accountability plays a role in adoption.
- The study challenges the assumption that military personnel will uncritically adopt AI, revealing a complex, multidimensional view of trust among senior officers.
Strategic Relevance
This research offers crucial insights into the human dimension of AI integration in military strategic decision-making. Understanding the factors that build or erode trust among senior officers is vital for developing effective AI systems, shaping military doctrine, and ensuring responsible and ethical deployment of AI in future conflicts. It directly informs policy on human-machine teaming and AI governance.