Understanding ‘AI Psychosis’: The Emerging Concern Over Mental Health and AI Chatbots

The term ‘AI psychosis’ has recently gained attention on social media as users describe experiences of losing touch with reality after prolonged interactions with AI chatbots like ChatGPT. These experiences reportedly include false beliefs, delusions of grandeur, and paranoid feelings following extended conversations with AI, often sought for low-cost therapy or professional advice. Although not clinically defined, the term is used informally to describe problematic online behavior akin to phenomena like ‘brain rot’ or ‘doomscrolling.’

This emerging concern coincides with the rapid growth of AI chatbots. ChatGPT alone, launched in 2022, is nearing 700 million weekly users. With this explosive adoption, mental health experts are sounding alarms about the potentially harmful psychological effects of intensive AI engagement. Vaile Wright, senior director for healthcare innovation at the American Psychological Association (APA), highlighted the lack of empirical data and stressed the need to study these anecdotal reports more rigorously.

The APA is responding proactively by forming an expert panel to investigate AI chatbot use in therapeutic settings and its effects on mental health. The panel aims to publish research and guidance within a few months, focusing on mitigating any harms arising from AI chatbot interactions. Addressing this issue is crucial given AI’s growing role in mental health support and daily information access.

Psychosis itself is a medically recognized condition characterized by symptoms like delusions, hallucinations, and disorganized thinking, usually linked to factors such as drug use, trauma, or underlying psychiatric conditions like schizophrenia. ‘AI psychosis,’ however, remains an informal term describing a pattern of user experiences that mirror some aspects of psychosis, without current clinical validation.

As AI technology continues to evolve and become more integrated into everyday life, understanding and addressing the psychological impact of prolonged AI interactions will be important. Mental health professionals and technology developers must collaborate to ensure AI tools are safe, supportive, and do not inadvertently contribute to mental health challenges.