Artificial intelligence (AI) psychosis: mechanisms, clinical risks and safety considerations in generative AI chatbots
Open Access
- 11 June 2026
- journal article
- Published by Royal College of Psychiatrists in BJPsych Open
- Vol. 12 (4) , e160
Abstract
Summary: As generative artificial intelligence chatbots become embedded in everyday life, concerns about their psychological risks are growing. Emerging reports describe cases of artificial intelligence-induced or -associated psychosis (hereafter artificial intelligence (AI) psychosis) in which intensive chatbot use is associated with delusional thinking patterns. This paper proposes a provisional mechanism wherein baseline user vulnerabilities and engagement patterns interact with generative artificial intelligence characteristics, such as sycophancy and hallucination, contributing to delusional ideation. It subsequently outlines clinical, design and regulatory strategies that may help mitigate risks.Keywords
This publication has 11 references indexed in Scilit:
- Sycophantic AI decreases prosocial intentions and promotes dependenceScience, 2026
- Technological folie à deux: feedback loops between AI chatbots and mental healthNature Mental Health, 2026
- Potentially Harmful Consequences of Artificial Intelligence ( AI ) Chatbot Use Among Patients With Mental Illness: Early Data From a Large Psychiatric Service SystemActa Psychiatrica Scandinavica, 2026
- User experience and safety of generative AI-based mental health chatbots: Scoping review protocolPLOS ONE, 2026
- Do generative AI chatbots increase psychosis risk?World Psychiatry, 2026
- Delusional Experiences Emerging From AI Chatbot Interactions or “AI Psychosis”JMIR Mental Health, 2025
- “You're Not Crazy”: A Case of New-onset AI-associated Psychosis2025
- AI driven psychosis and suicide are on the rise, but what happens if we turn the chatbots off?BMJ, 2025
- Special Report: AI-Induced Psychosis: A New Frontier in Mental HealthPsychiatric News, 2025
- AI hallucination: towards a comprehensive classification of distorted information in artificial intelligence-generated contentHumanities and Social Sciences Communications, 2024