An Empirical Study Exploring Difference in Trust of Perceived Human and Intelligent System Partners

Abstract

Intelligent systems are increasingly relied on as partners used to make decisions in business contexts. With advances in artificial intelligence technology and system interfaces, it is increasingly difficult to distinguish these system partners from their human counterparts. Understanding the role of perceived humanness and its impact on trust in these situations is important as trust is widely recognized as critical to system adoption and effective collaboration. We conducted an exploratory study involving individuals collaborating with an intelligent system partner to make several critical decisions. Measured trust levels and survey responses were analyzed. Results suggest that greater trust is experienced when the partner is perceived to be human. Additionally, the attribution of partners possessing expert knowledge drove perceptions of humanness. Partners viewed to adhere to strict syntactical requirements, displaying quick response times, having unnatural conversational tone, and unrealistic availability contributed to perceptions of partners being machine-like.