Acceptance of Artificial-Intelligence Chatbots for Support with Depressive Symptoms: The Predictive Primacy of Perceived Usefulness Over Well-Being

Abstract

Artificial-intelligence (AI) chatbots are increasingly proposed as low-threshold instruments for supporting peoplewith depressive symptoms, yet it remains unclear whether their acceptance is governed by users’ psychological needor by instrumental appraisal. Drawing on the Technology Acceptance Model, the present study reports originalempirical data from a cross-sectional online survey that examined whether the acceptance of AI chatbots fordepressive support is predicted by perceived usefulness, by psychological well-being, and by prior therapyexperience. A non-probabilistic, self-selected sample of 99 adults (M = 36.56 years, SD = 10.36) was recruitedthrough topic-relevant communities on the social-media platform Reddit and completed three-item Likert scalesassessing well-being, perceived usefulness, and acceptance. Perceived usefulness correlated strongly withacceptance, r = .77, p < .001, whereas well-being was unrelated to acceptance, r = -.00, p = .989. A multipleregression explained a substantial proportion of the variance in acceptance, R² = .60, adjusted R² = .59, F(2, 96) =72.45, p < .001, with perceived usefulness emerging as the sole significant predictor, β = .78, p < .001, while well-being did not contribute, β = .05, p = .55. Acceptance did not differ between participants with and without priortherapy experience, t(97) = 0.32, p = .748. These findings indicate that the acceptance of AI chatbots for depressivesupport follows an instrumental rather than a need-based logic: it is driven almost entirely by perceived usefulness,whereas current well-being and prior therapy experience play no discernible role. Implications for the design andframing of digital mental-health tools are discussed.