Authentically Imperfect: Pertinacious Images as a Strategic Approach of Stronger Mental Health Support AI Chatbots Adoption and Engagement

ABSTRACT AI chatbots offer accessible support in addressing the growing mental health crisis, yet users often hesitate to adopt these tools due to authenticity concerns, particularly when chatbots attempt to mimic empathy despite their nonhuman nature. Drawing on schema theory, expectancy violation theory, and inoculation theory, this study conceptualizes and experimentally tests the pertinacious image as an image-building strategy that acknowledges AI’s inherent limitations while emphasizing its unique strengths. By promoting an intentionally imperfect chatbot persona, this approach reduces user counterarguing and enhances perceived authenticity and engagement. The findings offer both theoretical and practical implications for the effective and ethical promotion of emotional support AI chatbots.