Study Links Generative AI Use for Health Queries to Higher Anxiety and Depression in Young Adults

The study revealed that individuals using generative AI tools for health questions exhibited a 52% higher likelihood of clinically significant anxiety and a 46% higher likelihood of depression compared to those who did not use these tools. However, the findings indicate an association rather than a direct causation; they do not imply that AI usage leads to anxiety or depression, nor that those with these mental health issues are necessarily more inclined to seek health information from AI.

Zhai expressed interest in the results, noting that discussions around AI in education typically focus on its application for assignments and productivity. His research aims to explore the intersection of AI, mental health, and technological approaches to identifying mental health risks. He highlighted that young people are increasingly turning to AI during times of worry or vulnerability to better understand their health symptoms.

The study found that among AI users, approximately 52% screened positive for anxiety, while about 43% of non-users did. For depression, the figures were about 47% for AI users and 38% for those who did not use AI for health inquiries. These correlations persisted even after adjusting for factors like age, race, biological sex, and socioeconomic status.

Zhai noted that repeated reliance on AI during health concerns might create a cycle of uncertainty. Initially, a user may ask an AI a question for reassurance, which could lead to further inquiries and, potentially, increased anxiety or health concerns. The interactive nature of generative AI differs from traditional internet searches since it provides synthesized answers and can respond to follow-up questions directly.

This conversational approach may make AI responses feel more persuasive, even if they include incorrect or misleading information. Zhai warned that young individuals might struggle to recognize inaccuracies when they are presented within seemingly plausible answers.

The study gathered data between 2023 and 2024, a time when generative AI was just starting to gain widespread use. As generative AI's popularity has grown, the trends observed in the study may reflect evolving behaviors among young people as they engage with AI for health information.

Zhai emphasized the need for further research to understand the dynamics of the relationship between AI usage, anxiety, and depressive symptoms. While some individuals seeking health information from AI may be in search of support, frequent reassurance-seeking could intensify their worries.

In summary, Zhai asserted that the focus should be on how individuals use AI. He encouraged young users to reflect on their emotional responses after interacting with generative AI regarding health inquiries. Awareness of how such interactions affect their mental state is a part of developing AI literacy.

Mental health professionals should inquire about clients' experiences with generative AI to aid in evaluating the information they receive. Furthermore, universities could integrate guidance on using AI responsibly as part of broader educational initiatives aimed at fostering AI literacy, ensuring students can identify reliable health information and understand the constraints of AI-generated responses. Zhai concluded that the issue lies not with generative AI itself, but rather how it is utilized.