Abstract

As AI bots become increasingly integrated into everyday interactions, their design—particularly the use of anthropomorphic features—plays a critical role in shaping user perceptions. This study investigates the impact of anthropomorphic cues displayed by AI bots on distinct components of trust: perceived benevolence, integrity, competence, and behavioural trust, as reflected in the intention to follow the advice, across various industry contexts. Using a 2×2 between-subjects experimental design, we manipulate anthropomorphic design (low vs. high) and industry context (high-stakes vs. low-stakes) to examine user responses in simulated healthcare and retail scenarios. We hypothesise that anthropomorphism enhances engagement in low-stakes contexts but diminishes trust and perceived appropriateness in high-stakes settings. The findings of this research-in-progress paper aim to inform ethical and effective AI design strategies tailored to specific contexts and user demographics.

Download

Share

COinS