Automated intelligent agents are they trusted members of military teams?

Abstract

Technological advances and increased operational challenges have led to the introduction of automated agents into military teams. Although these new combined teams have many advantages, it is possible that the interactions between members of these new human-automation teams may adversely impact mission accomplishment. This study investigates the similarities and differences between human-human teams and human-automation teams with respect to team communications, efficacy, and trust. Thirty-six participants were formed into twelve three-person teams. A confederate served as the fourth member for all twelve teams. In the human-human team condition, the confederate was present in the same room as the other three team members. In the human-automation team condition, the confederate was located in a separate room and the other three team members were told that their fourth team member was an automated intelligent agent. All teams played a computer-based team firefighting game (C3Fire). The order of presentation of the two trials (human-human vs. human-automation) was counterbalanced. The results of this study indicate there is a significant difference in the nature of the communication between these two types of teams. Additionally, the presence of an automated agent changes the nature of trust and team efficacy. These findings demonstrate the need to consider the unintended impact of including automated agents on team dynamics in military environments and other complex and dynamic systems.