Trust in artificial intelligence, trust in engineers, and news media: Factors shaping public perceptions of autonomous drones through UTAUT2

Traditional trust in artificial intelligence (AI) scholarship has assumed trust as an attitude placed exclusively in the AI system. We offer a more articulated view of trust in AI, extending the extended Unified Theory of Acceptance and Use of Technology (UTAUT2) model by integrating news media attention, trust in AI system, trust in AI engineers, and attitude to examine factors predicting public's intention to use autonomous passenger drones. Based on an online survey of adult Singaporeans (N = 1002), results from hierarchical OLS regression analyses demonstrated the direct effects of news media attention, trust in AI system, and attitude, alongside several UTAUT2 constructs, on use intention. Results of subsequent mediation analyses using structural equation modelling between trust in AI system, trust in AI engineers and use intention through attitude were consistent with a partial and full mediation effect, respectively. Notably, both trust constructs were influenced by news media attention in comparable magnitudes. In this sense, trust placed in AI system can be seen separately from that placed in engineers responsible for AI development. Furthermore, we found support for including attitude in the UTAUT2 main model for emerging technologies in which information and practice remain largely inaccessible to the public. Theoretical and practical implications are discussed.