AI-integrated chatbots are common tools for work and daily life, but clinicians warn that repeated interactions can reinforce delusions in vulnerable people. Anthony Cesar Duncan of Missouri says his 2023 ChatGPT use led to conspiracy-driven beliefs that cost him his apartment, car, business network and full-time job and resulted in an involuntary psychiatric hold (often called a '5150' in some areas). Experts stress that 'AI psychosis' is not an official diagnosis but is a growing phenomenon and recommend monitoring usage and applying gradual reductions rather than abrupt cuts. They also call for safeguards as AI becomes more embedded in daily life.
Man Says 'AI Psychosis' From ChatGPT Cost Him Job, Home and Sanity — Clinicians Warn

Artificial intelligence is increasingly woven into everyday life, helping with tasks from work and education to personal organization. But mental health experts are raising concerns about a troubling pattern some call 'AI psychosis' — repeated interactions with chatbots that may reinforce delusional thinking in vulnerable users.
Case Study: One Man's Experience
Anthony Cesar Duncan of Missouri told NewsNation that after using ChatGPT for business in 2023, his conversations shifted from practical help to conspiracy-driven exchanges. He says the chatbot repeatedly affirmed those ideas until they became fixed beliefs.
'That spiraled into talking to it about conspiracy theories, and they became delusions as soon as the chatbot would triple down, affirming that they were real, presenting it as facts,' Duncan said.
Duncan says the episode had devastating consequences: 'I lost my apartment, I lost my car, my business network, my full-time job and ended up 5150'd in a psych ward,' he said. (A '5150' commonly refers to an involuntary psychiatric hold in some U.S. jurisdictions.)
'I even thought that I had put like 10% of my soul into AI. It's just wild to think about,' Duncan added.
Expert Perspective
The National Institute of Mental Health defines psychosis as a set of symptoms that can cause a person to lose touch with reality. Clinicians caution that while 'AI psychosis' is not a formal medical diagnosis, repeated validating interactions with AI can deepen unhealthy thought patterns in susceptible individuals.
Dr. Matthew Leahy, a clinical psychologist, said, 'AI bots are validating machines … and before you know it, conversations start to get deeper and deeper and go down different rabbit holes. Situations like his are really manifesting and coming up. We're seeing it a lot more. We're not diagnosing it, but it's a phenomenon that's growing.'
Recommendations and Next Steps
Clinicians advise monitoring how much time you spend interacting with AI and watching for signs of increasing dependence, such as late-night conversations that interfere with sleep or daily life. Rather than cutting off AI use abruptly, mental health professionals often recommend a gradual reduction and professional support when needed.
Experts also urge policymakers and developers to consider safeguards as AI becomes more embedded in daily life to mitigate unintended consequences.
Source: NewsNation. Copyright 2026 Nexstar Media, Inc.
Help us improve.
























