AI chatbot-associated psychiatric symptoms: a narrative review of publicly available reports and media accounts

Abstract

Abstract Background There are a growing number of cases where direct use of AI has been associated with either new or worsening psychiatric symptoms (colloquially labelled AI psychosis). Due to the novelty of the phenomenon, current literature is limited on the impacts of sustained, emotionally charged interactions with AI chatbots. This narrative review of publicly available reports and media accounts aims to address this gap by identifying reports of AI-associated psychiatric symptoms. Methods Given the novelty of the topic, reports published from inception to March 1, 2026, were identified from Google News and Google Scholar. Search terms included: “artificial intelligence,” “AI chatbot,” “ChatGPT,” “Character.AI,” and “large language model,” combined with terms such as “psychosis,” “delusion,” “depression,” “anxiety,” “suicide,” and “spiritual.” These terms were selected to focus on reports that describe psychiatric symptoms associated with AI use, while excluding general articles on AI ethics, technology, or public opinion. Two authors conducted an independent analysis of the results, and all authors reached a consensus on the publications to include in this review. While websites such as Reddit contain over 100 different threads of cases dealing with AI chatbot-associated psychiatric symptoms, they were excluded due to a lack of information. Results Nineteen cases were identified describing users who developed a sustained emotional reliance on AI chatbots. Across cases, this developed regardless of the original reason for chatbot use and was associated with the emergence of psychiatric symptoms such as anxiety, depressive symptoms, delusional beliefs, and suicidal ideation. In many cases where the user expressed suicidal ideation, chatbots responses reportedly lacked appropriate crisis intervention and, in some cases, reinforced the users’ beliefs. Conclusion AI chatbot-associated psychiatric symptoms are an emerging concern with new cases increasingly being reported. This review explores such cases while also serving as a call to action for more research and discussion on how to address the risks of AI use.