Trust repair in human-agent teams: the effectiveness of explanations and expressing regret

Abstract The role of intelligent agents becomes more social as they are expected to act in direct interaction, involvement and/or interdependency with humans and other artificial entities, Human-Agent Teams (HAT). highly interdependent dynamic nature teamwork demands correctly calibrated trust among team members. Trust violations an inevitable aspect the cycle since repairing damaged proves be difficult than building initially, effective repair strategies needed ensure durable successful performance. aim this study was explore effectiveness different from agent by measuring development human advice taking a Teaming task. Data for were obtained using task environment resembling first-person shooter game. Participants carried out mission collaboration their member. A violation provoked when failed detect approaching enemy. After this, offered one four strategies, composed apology components explanation expression regret (either alone, both or neither). Our results indicated that expressing crucial repair. declined due agent, only significantly recovered included apology. This effect stronger added. In context, most its attempt rebuilding it provided affective, informational. Finally, implications our findings design discussed.