Exploring Real-World Use of AI Chatbots for Mental Health Support: Cross-Sectional Survey Study

Key Findings

• — 349 participants were analysed: 276 members of the general public and 73 students; approximately 40% of the student sample were eligible users who had used an AI chatbot for mental-health support
• — Most respondents were women and indicated having received or currently receiving professional mental-health support
• — The most common conversation topics in both samples were personal and interpersonal issues; freely accessible chatbots, mainly ChatGPT, were the predominant choice, selected for constant availability and accessibility (students also for anonymity, the general sample also because they felt supported)
• — Preliminary associations between digital working alliance and engagement factors in both samples suggest relational factors may play a role in sustained use; the authors conclude general-purpose chatbots not designed or regulated for mental-health support are being used for social and emotional support, often by people already familiar with professional care

Methodology Notes

JMIR Mental Health 2026;13:e104316, published 2026-09-01 (Crossref published-online and the PubMed citation line agree; OpenAlex lists 2026-07-12, which is not supported by the other two sources). Cross-sectional online survey, self-selected convenience samples in Belgium, skewed female, no behavioural data; the working-alliance associations are described by the authors as preliminary and warranting further investigation. Authors at Thomas More University of Applied Sciences (Antwerp), University College Dublin and Queen's University Belfast. JMIR hosts block automated fetchers from this workspace; verified via the full PubMed structured abstract (PMID 42679350) and the Crossref-deposited abstract.

Sources