Trust Propagation in Multi-Operator Multi-Autonomy (MOMA) Teams: Friends Versus Strangers

Trust in automation plays a pivotal role in multi-operator, multi-agent (MOMA) teams, particularly in high-stakes domains such as defense operations or disaster response. This study examines how trust in autonomous systems propagates between human teammates and how interpersonal familiarity modulates that process. In a simulated threat-detection task with N = 64 participants, friend and stranger pairs collaborated across 15 sessions with two drones of varying reliability. Participants responded to drone suggestions but retained final decision-making authority. Trust ratings were collected for directly and indirectly experienced drones. Our results confirm that trust propagation occurred in both friend and stranger teams. Contrary to expectation, the magnitude of propagation did not significantly differ by familiarity. However, trust in the human teammate was substantially higher among friend dyads. This finding reveals a novel dissociation: while familiarity strengthens interpersonal trust, it does not necessarily amplify trust propagation toward automation. These results highlight the distinct and interactive roles of social and informational trust in shaping human-autonomy teaming.