Private Yet Social: How LLM Chatbots Support and Challenge Eating Disorder Recovery
Ryuhaerang Choi, Taehan Kim, Subin Park, Jennifer G Kim, Sung-Ju Lee
TL;DR
Eating disorders demand sustained support beyond clinical care, and LLM-based chatbots offer immediate, private assistance but introduce safety and reliability risks. This study deploys WellnessBot, a technology probe, in a 10-day field study with 26 ED participants to examine how storytelling, personalized ED-context understanding, and long-term memory can empower recovery while highlighting potential harms from uncritical trust and harmful outputs. The findings show that WellnessBot can create a private, socially present space that fosters self-reflection and goal-oriented engagement, yet it can also produce inappropriate guidance that participants may not challenge. The authors discuss design implications for human-LLM collaboration, critical-thinking prompts, and context-aware safeguards to balance empowerment with safety in ED care and other mental health contexts.
Abstract
Eating disorders (ED) are complex mental health conditions that require long-term management and support. Recent advancements in large language model (LLM)-based chatbots offer the potential to assist individuals in receiving immediate support. Yet, concerns remain about their reliability and safety in sensitive contexts such as ED. We explore the opportunities and potential harms of using LLM-based chatbots for ED recovery. We observe the interactions between 26 participants with ED and an LLM-based chatbot, WellnessBot, designed to support ED recovery, over 10 days. We discovered that our participants have felt empowered in recovery by discussing ED-related stories with the chatbot, which served as a personal yet social avenue. However, we also identified harmful chatbot responses, especially concerning individuals with ED, that went unnoticed partly due to participants' unquestioning trust in the chatbot's reliability. Based on these findings, we provide design implications for safe and effective LLM-based interventions in ED management.