Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Conditional Companion: Lived Experiences of People with Mental Health Disorders Using LLMs

Aditya Kumar Purohit, Hendrik Heuer

TL;DR

This study investigates how people with mental health challenges use general-purpose LLMs for everyday support in the UK. Through 20 semi-structured interviews and reflexive thematic analysis, it reveals a mode of use characterized by immediacy, non-judgmental disclosure, self-paced interaction, sensemaking, and relational engagement, alongside boundary conditions that limit LLMs to mild-to-moderate distress and not crises or complex social scenarios. The findings articulate design and governance directions—safety with escalation pathways, accessible and structured interactions, personalization with memory controls, integration into broader care, and clinician collaboration modes—framing LLMs as supplementary tools within care ecosystems rather than replacements for human therapists. Practically, the work informs responsible AI design, regulatory considerations, and patient-centered approaches to balance accessibility with safety in high-stakes mental health contexts.

Abstract

Large Language Models (LLMs) are increasingly used for mental health support, yet little is known about how people with mental health challenges engage with them, how they evaluate their usefulness, and what design opportunities they envision. We conducted 20 semi-structured interviews with people in the UK who live with mental health conditions and have used LLMs for mental health support. Through reflexive thematic analysis, we found that participants engaged with LLMs in conditional and situational ways: for immediacy, the desire for non-judgement, self-paced disclosure, cognitive reframing, and relational engagement. Simultaneously, participants articulated clear boundaries informed by prior therapeutic experience: LLMs were effective for mild-to-moderate distress but inadequate for crises, trauma, and complex social-emotional situations. We contribute empirical insights into the lived use of LLMs for mental health, highlight boundary-setting as central to their safe role, and propose design and governance directions for embedding them responsibly within care ecosystem.

A Conditional Companion: Lived Experiences of People with Mental Health Disorders Using LLMs

TL;DR

This study investigates how people with mental health challenges use general-purpose LLMs for everyday support in the UK. Through 20 semi-structured interviews and reflexive thematic analysis, it reveals a mode of use characterized by immediacy, non-judgmental disclosure, self-paced interaction, sensemaking, and relational engagement, alongside boundary conditions that limit LLMs to mild-to-moderate distress and not crises or complex social scenarios. The findings articulate design and governance directions—safety with escalation pathways, accessible and structured interactions, personalization with memory controls, integration into broader care, and clinician collaboration modes—framing LLMs as supplementary tools within care ecosystems rather than replacements for human therapists. Practically, the work informs responsible AI design, regulatory considerations, and patient-centered approaches to balance accessibility with safety in high-stakes mental health contexts.

Abstract

Large Language Models (LLMs) are increasingly used for mental health support, yet little is known about how people with mental health challenges engage with them, how they evaluate their usefulness, and what design opportunities they envision. We conducted 20 semi-structured interviews with people in the UK who live with mental health conditions and have used LLMs for mental health support. Through reflexive thematic analysis, we found that participants engaged with LLMs in conditional and situational ways: for immediacy, the desire for non-judgement, self-paced disclosure, cognitive reframing, and relational engagement. Simultaneously, participants articulated clear boundaries informed by prior therapeutic experience: LLMs were effective for mild-to-moderate distress but inadequate for crises, trauma, and complex social-emotional situations. We contribute empirical insights into the lived use of LLMs for mental health, highlight boundary-setting as central to their safe role, and propose design and governance directions for embedding them responsibly within care ecosystem.
Paper Structure (41 sections, 1 figure, 4 tables)

This paper contains 41 sections, 1 figure, 4 tables.

Table of Contents

  1. Introduction
  2. Scope
  3. Background and Related Work
  4. Chatbots and Mental Health Self-care
  5. LLMs in Mental Health Care
  6. Fine-tuning LLMs for mental health.
  7. LLMs for screening and risk assessment.
  8. LLMs as supportive counseling agents.
  9. Method
  10. Ethical Consideration
  11. Participant Screening and Survey
  12. Interviews
  13. Participants
  14. Interview Protocol
  15. Qualitative Analysis
  16. ...and 26 more sections

Figures (1)

  • Figure 1: Participants' articulation of the boundary of LLM-mediated mental health support across a continuum of distress. Everyday regulation and mild-to-moderate concerns were often seen as appropriate for LLMs, while more complex or severe needs were consistently placed beyond this boundary and viewed as requiring professional care.