Can AI be a trusted counselor? How AI counseling builds psychological empowerment and compliance intention

This study examines how perceptions of an AI counselor influence trust, health empowerment, and compliance intention within an AI digital health intervention. Using covariance-based structural equation modeling (CB-SEM), the research reveals that perceived transparency, ethical perception, and empathy are the primary predictors of trust, whereas perceived intelligence and fairness are statistically non-significant. These findings suggest that the success of AI-mediated support depends on relational qualities rather than technological sophistication alone. The results demonstrate that trust facilitates the co-creation of health value by driving both compliance intention and psychological empowerment. Notably, the study identifies an Uncanny Empathy Paradox, where significantly high levels of simulated emotional warmth negatively influence trust. Furthermore, mediation analysis confirms a serial pathway where trust fosters health empowerment, which subsequently drives compliance. By positioning empowerment as a critical mediator, this research provides a theoretical foundation for managing AI as a supplemental resource, rather than a substitute for professional clinical care. The findings emphasize that responsible service delivery requires clear ethical boundaries and transparent communication to foster autonomous health management.