Answers are grounded in this paper and cite their sources.
Despite increasing efforts of legislatures to prohibit specialized AI-driven apps from providing therapeutic decision-making, general-purpose Large Language Models (LLMs) are increasingly being used as alternative tools for mental health support. In this paper, we examine the increased adoption of AI for mental health through the lens of proponents’ and opponents’ discussions on public forums. Through a qualitative analysis of 67 Reddit discussion threads (totaling 2948 comments), we offer an in-depth investigation of the various contexts in which users rely on (or fail to rely on) AI for mental health support, as well as an analysis of the main discussions (including positions for and against the use of AI for therapy) within the community. By critically reflecting on our findings, we highlight ethical considerations and discuss how research, design and regulation of AI tools can help improve mental health support for users with complicated needs while mitigating potential harms.
No takes yet. Share an insight, caveat, or question.
Lee et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: