Why Some Seek AI, Others Seek Therapists: Mental Health in the Age of Generative AI

TL;DR

The paper investigates how people decide between generative AI and human therapists for mental health by applying a cross-modality Health Belief Model across university and national adult samples. It uses RM-ANOVA and LASSO to identify belief-based predictors of intention to use each modality and to examine cross-modality effects. Key findings show therapists are valued for emotional and relational aspects, while GAI is favored for accessibility and affordability; however, emotional resonance and personalization are the strongest predictors of adoption, with divergent patterns of substitution versus complementarity across populations. The study offers design implications for trustworthy, emotionally credible GenAI tools and highlights the need for pathway-oriented, literacy-centered approaches to integrate AI into broader mental health ecosystems.

Abstract

As generative artificial intelligence (GAI) enters the mental health landscape, questions arise about how individuals weigh AI tools against human therapists. Drawing on the Health Belief Model (HBM), this study examined belief-based predictors of intention to use GAI and therapists across two populations: a university sample (N = 1,155) and a nationally representative adult sample (N = 651). Using repeated-measures ANOVA and LASSO regression, we found that therapists were consistently valued for emotional, relational, and personalization benefits, while GAI was favored for accessibility and affordability. Yet structural advantages alone did not predict adoption; emotional benefit and personalization emerged as decisive factors. Adoption patterns diverged across groups: students treated GAI as a complement, whereas national adults approached it as a substitute. Concerns about privacy and reliability constrained GAI use in both groups. These findings extend HBM to multi-modality contexts and highlight design implications for trustworthy, emotionally resonant digital mental health tools.

Figures (4)

• Figure 1: Comparison of a single-modality versus multi-modality Health Belief Model (HBM) approach for explaining intentions to use Generative AI (GAI) or a human therapist. The single-modality panel models benefits and barriers separately for GAI and for therapists with direct effects on each intention. The multi-modality panel adds cross-modality influences, where benefits and barriers for both GAI and therapists jointly increase or decrease intention to use GAI.
• Figure 2: Profiles of the top three perceived benefits and barriers for Generative AI (GAI) and human therapists, aggregated across the combined university and national samples ($N=1{,}806$). The figure shows that GAI is most valued for accessibility, affordability, and educational benefits, but is limited by privacy, reliability, and technology literacy concerns. Human therapists are valued for emotional benefit, personalization, and practical benefit, but face barriers of privacy, reliability, and stigma.
• Figure 3: Summary of influential perceived benefits and barriers predicting intention to use Generative AI (GAI) or a human therapist in the university sample ($N=1{,}155$). For visibility, only the two largest coefficients in each category are shown. Full results are reported in Table \ref{['tab:lasso_results']}.
• Figure 4: Summary of influential perceived benefits and barriers predicting intention to use Generative AI (GAI) or a human therapist in the national sample ($N=651$). For visibility, only the two largest coefficients in each category are displayed. Full results are reported in Table \ref{['tab:lasso_results']}.