
Exploring Trust and Autonomy: How Information Affects Human-Agent Teaming Performance
Proc. SPIE 13477, Unmanned Systems Technology XXVII, 2025
Best Paper Award

Abstract
Human-agent teams encounter many challenges, including a lack of common understanding and unbalanced reliance and trust among teammates. When humans lack trust in agents, they may be less inclined to collaborate with them, leading to inefficient performance due to an imbalance in workload. This study explores how different information availability conditions affect reliance and trust between a human and two virtual autonomous agents as they complete a collaborative search task. As more information became available, participants were expected to continue to rely on the agents, while their trust increased. However, the results were not as straightforward. With the different information availability conditions came different patterns of trust and reliance.