Impact of AI recommendation styles on carbon management technicians' trust: Dual process of cognitive effort and emotion

Artificial intelligence (AI), as an advanced technological innovation, possesses considerable promise for improving energy efficiency and reducing emissions in industrial environments. Although AI plays a crucial role in industrial decarbonization, research reveals that many AI-dependent carbon reduction prospects remain untapped. This study examines the innovative correlation between the presentation styles of industrial AI recommendations and the trust that carbon management technicians have in AI from an ergonomic standpoint. This study combines the Stimulus-Organism-Response theory with the Elaboration Likelihood Model. The findings from three experiments reveal that precise and informative styles of AI recommendations enhance trust among technicians. Furthermore, we present mediating evidence with advanced eye-tracking devices, illustrating that cognitive effort strengthens the positive effect of precise abatement recommendations on trust toward AI (Study 2). Further research indicates that the positive effect of precise or informative AI-generated abatement recommendations on trust is amplified when technicians experience more positive emotions (Study 3). Overall, this study identifies distinct mechanisms through which the presentation styles of industrial AI recommendations can affect technicians' trust. These findings hold practical significance for developers of industrial AI systems and carbon management technicians who strategically utilize AI-generated information.