Exploring the Design of a LLM-Based AI Assistant for Mindfulness Practice With Older Adults
Lucy McCarren, Ulrika Eriksson, Laura Ortiz Mengual, Sanna Kuoppamäki
International Conference on Human Factors in Computing Systems April 13, 2026 DOI: 10.1145/3772318.3790734 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Qualitative study Peer reviewed |
|---|---|
| Sample size | 16 |
| Population | Older adults |
| Topics | Meditation |
| Key findings | Older adults experienced tensions between adaptivity and autonomy, supportive versus intrusive engagement, and AI emotional support versus human connection. Design considerations should address these tensions to create LLM-based mindfulness tools that respect older adults' autonomy and self-regulation. |
Abstract
Large Language Models (LLMs) are increasingly integrated into mental health and well-being technologies, yet little is known about how they are perceived by older adults or how they should be designed to meet later-life needs. Mindfulness technologies, often promoted as tools for healthy ageing, provide a useful context for exploring these questions. We conducted participatory workshops with sixteen older adults using LugnAI, a prototype LLM-based system for guided mindfulness practice. Participants reflected on their experiences with AI-guided mindfulness and contributed design preferences for future systems. Analysis revealed tensions between adaptivity and autonomy, supportive versus intrusive engagement strategies, and AI-enabled emotional support versus the preservation of human connection and self-regulation practices. Based on these findings, we provide concrete design considerations for LLM-based mindfulness technologies that are sensitive to the socioaffective realities of ageing. While situated in mindfulness, the insights extend to broader applications of LLMs in supporting older adults’ well-being.