<b>Does the “Assistant” Heuristic Really Work? Analyzing the Effect of Perceived Hierarchy of AI through Assigned Job Titles on the Adoption of Conversational AI-based Services</b>

Recent marketing trends involve companies using low-status job titles, such as "assistant" (e.g., Google Home Assistant), to label conversational AI agents. This strategy aims to activate the “helper” heuristic and improve people's willingness to use these AI agents. However, this paper — which includes two experiments (N=307, N=300), one partial least squares structural equation modeling analysis (N=309), and one set of interviews (N=15) — reveals the opposite effect. The current research tests four alternative serial mediation pathways involving perceived humanlikeness: perceived warmth, perceived competence, perceived control, and perceived risks. Across the four studies, only the serial mediation pathway through perceived humanlikeness and perceived competence was consistently supported. Our final model indicates that in conversational AI settings, increased perceptions of hierarchical status enhance perceived humanlikeness. This, in turn, boosts perceived competence, which subsequently increases the willingness to use the AI.