A U.S.-based psychiatrist, Dr. Keith Sakata, has raised alarm over what he calls “AI psychosis,” claiming he has treated 12 patients in 2025 who were hospitalized after losing touch with reality due to interactions with artificial intelligence.
The statement, shared on X (formerly Twitter), has sparked debate among mental health professionals and AI researchers. His tweet is gaining more and more upvotes.
According to him, these patients were not ordinary tech users but individuals with pre-existing vulnerabilities to psychosis or delusional thinking. He stresses that AI does not directly cause psychosis; rather, it can act as a trigger, amplifying pre-existing mental health issues. “I’m seeing the same pattern online,” he wrote, noting that intense engagement with AI chatbots or immersive AI-generated worlds can blur the line between reality and fiction for at-risk users.
Psychosis is a mental health condition where a person loses touch with reality. It usually involves: Hallucinations, Delusions, Disorganized thinking & Distorted Thinking
The phenomenon, sometimes referred to as “chatbot psychosis”, is not yet an official medical diagnosis, but it is drawing attention from clinicians worldwide. Mental health experts warn that as AI tools become more conversational, personalized, and emotionally responsive, they may create dependency or reinforce delusional beliefs in certain individuals.
While most AI users interact without harm, experts say it’s critical to monitor vulnerable populations especially those with schizophrenia, bipolar disorder, or other psychiatric conditions. Researchers are calling for ethical guidelines, user safety mechanisms, and public awareness campaigns to minimize potential risks. Dr. Sakata’s observations highlights a broader challenge in the AI era.
Read the tweet: X (Formerly Twitter) handle of Dr. Sakata
Disclaimer: For more medical technology and digital heath updates, follow #MedTechwithGPHSA. This content is for informational purposes only and is not a substitute for professional medical advice.
