The opportunities and risks of large language models in mental health

TL;DR

This paper addresses the global rise in mental health concerns and the insufficient access to care by evaluating how large language models (LLMs) can support mental health education, assessment, and intervention. It distinguishes general-purpose LLMs from domain-specific ones and reviews evidence across education, screening/diagnosis, and therapeutic chatbots, highlighting both potential benefits and limitations. Key contributions include synthesizing findings on accuracy, reliability, ethics, equity, and the role of human oversight, and proposing risk-mitigation strategies such as domain-focused fine-tuning, transparent reporting, and stakeholder involvement. The work suggests that responsible development and deployment of mental health LLMs—centered on equity, evidence-based practice, and confidentiality—could expand access to mental health information and services globally, though substantial research remains to ensure safety and effectiveness across diverse populations.

Abstract

Global rates of mental health concerns are rising, and there is increasing realization that existing models of mental health care will not adequately expand to meet the demand. With the emergence of large language models (LLMs) has come great optimism regarding their promise to create novel, large-scale solutions to support mental health. Despite their nascence, LLMs have already been applied to mental health related tasks. In this paper, we summarize the extant literature on efforts to use LLMs to provide mental health education, assessment, and intervention and highlight key opportunities for positive impact in each area. We then highlight risks associated with LLMs' application to mental health and encourage the adoption of strategies to mitigate these risks. The urgent need for mental health support must be balanced with responsible development, testing, and deployment of mental health LLMs. It is especially critical to ensure that mental health LLMs are fine-tuned for mental health, enhance mental health equity, and adhere to ethical standards and that people, including those with lived experience with mental health concerns, are involved in all stages from development through deployment. Prioritizing these efforts will minimize potential harms to mental health and maximize the likelihood that LLMs will positively impact mental health globally.

Figures (5)

• Figure 1: Potential opportunities for LLMs in mental health education. CBT: cognitive behavioral therapy; EST: empirically supported treatment; LLM: large language model.
• Figure 2: Potential opportunities for LLMs in mental health assessment. LLM: large language model.
• Figure 3: Potential opportunities for LLMs in mental health intervention. CBT: cognitive behavioral therapy; EBP: evidence-based practice; LLM: large language model.
• Figure 4: Textbox1: Recommendations for responsible use of LLMs to support mental health.
• Figure 5: Examples of human involvement across all stages of LLM development through deployment and evaluation. LLM: large language model.