Abstract
OBJECTIVES: In some services involving social and emotional needs (such as psychological counseling), people seem to prefer human services over AI. Exploring the psychological mechanisms behind this preference can help increase people's acceptance of AI-based psychological counseling. This study aims to explore the differences in people's consultation intention for human versus AI across different counseling scenarios, as well as the role of trust (including cognitive and emotional trust) between humans and AI. METHODS: A total of 477 participants (297 in Study 1 and 180 in Study 2) were randomly assigned to different groups for counseling imagination tasks and then completed self-report questionnaires. RESULTS: The results of Study 1 demonstrated a significantly higher consultation intention towards human counselors in social emotional scenarios, while no significant preference was observed in cognitive analytical scenarios. Study 2 replicated the findings of Study 1, and further revealed that: (1) cognitive and affective trust played a multiple mediating role between counselor type and consultation intention in social emotional scenarios; (2) there existed a suppressing effect in the relationship model between cognitive trust, counselor type, and counseling intentions in cognitive analytical scenarios; (3) psychological counseling scenarios moderated the relationship between cognitive trust/counselor type and consultation intention. CONCLUSION: These findings offer practical guidance for the development of AI-driven psychological consultation products and carry theoretical implications for research pertaining to human-AI interaction.