AI-driven mental health therapy: A comparative study of emotional clarity and stigma reduction across journaling, human-guided therapy, and AI-guided therapy

Publication Type

Journal Article

Publication Date

8-2026

Abstract

As global demand for accessible mental health solutions grows, AI-driven therapy emerges as a promising alternative for addressing stress-related challenges. Anchored in social cognitive theory (SCT) and the transactional model of stress and coping (TMSC), this study investigates whether AI-mediated interventions can enhance emotional clarity and reduce mental health stigma compared to traditional approaches. Through a three-group experiment comparing AI-guided therapy, human-guided therapy, and journaling, we assess their differential impacts on emotional processing and stress relief. The findings reveal significant improvements across all interventions, with nuanced distinctions emerging. Both human and AI-guided therapies demonstrate superior efficacy over journaling in enhancing emotional clarity, while AI therapy uniquely excels in reducing perceived stigma associated with seeking help. Human and AI modalities show comparable efficacy in promoting adaptive coping strategies, alleviating stress, and increasing future help-seeking intentions, all of which significantly outperform journaling outcomes. Mediation analyses confirm that these effects operate through enhanced emotional clarity and reduced stigma perception. These results substantiate AI's potential as an effective therapeutic tool that reduces stigma and democratizes access to mental health support. The findings offer actionable insights for integrating AI into mental health ecosystems as a complementary or alternative intervention in stress management contexts.

Keywords

AI therapy, Emotional clarity, Mental health stigma, Coping strategies, Stress management

Discipline

Databases and Information Systems | Health Information Technology

Research Areas

Software and Cyber-Physical Systems

Publication

Decision Support Systems

Volume

207

Issue

C

First Page

1

Last Page

12

ISSN

0167-9236

Identifier

10.1016/j.dss.2026.114696

Publisher

Elsevier

Additional URL

https://doi.org/10.1016/j.dss.2026.114696

This document is currently not available here.

Share

COinS