Attitudes Toward LLM Use for Mental Health Support: A Comparative Study of India and the USA
Radhika Vikas Yadav
Department of Psychology
Faculty Supervisor: Gaurav Suri
Large language models (LLMs) are increasingly being used for mental health support, yet empirical understanding of public attitudes toward these tools remains limited. This paper presents two original surveys across Indian and American populations: the first examining usage patterns, perceived opportunities, and risks among current users, and the second exploring barriers to adoption among non-users. Alongside this, we offer a detailed review of a rapidly developing literature, tracing how LLMs function, where risks emerge, and what mitigations have been proposed. Drawing these strands together with primary data, we find that while LLM use for emotional support is already substantial, risk awareness varies considerably across populations. We conclude with evidence-based guidelines for responsible deployment, grounded in both the technical realities of these systems and the comparative patterns observed across our two samples.