Human-likeness perceptions in automated driving systems: exploring post-usage trust and continuance intention through TAM and automated social presence perspectives

Mind perception theory explains how people attribute human-like qualities to technology. Drawing on this theory, this study introduces perceived competence and warmth as key dimensions of human-likeness in automated driving systems (ADS). We propose a post-usage model for Level-3 ADS trust and adoption. It extends TAM by incorporating the two human-likeness dimensions, trust, and automated social presence (ASP; feeling of being socially accompanied by automation). We conducted a driving-simulator experiment to manipulate users' perceptions of competence and warmth. The proposed model was then validated using multilevel structural equation modelling with 280 experimental samples. Results show competence and warmth jointly enhance perceived ease of use, usefulness, and ASP, thereby promoting trust and continued usage. Notably, warmth receives greater user attention than competence. Moreover, post-usage trust exerts a stronger impact on continuance intention than original TAM pathways. Our findings inform the design of ADS that foster trust and continued adoption.