Conversational AI and Psychosis: A Proposed Neurocognitive Framework Involving Anthropomorphism, Predictive Coding, and Delusional Reinforcement
Pratik Kumar Banik, Heena Arshad, Aadil Ahmed Khan, Margarita Zhuzha, Inamullah Arshad, Amuksha Ravindra Shetty, Srushti Patil, Rahat Rehman
Cureus November 2, 2026 DOI: 10.7759/cureus.118999 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Narrative review Longitudinal Cross-sectional Case report Peer reviewed |
|---|---|
| Key findings | Argues that conversational AI use has been associated with psychotic-like experiences in cross-sectional surveys and implicated in onset or relapse of psychotic symptoms in an isolated case report, while proposed neurocognitive mechanisms remain hypothesis-generating rather than empirically demonstrated. The authors contend that conversational AI may also have supportive clinical uses, but should be approached with caution given potential psychological risks. |
Abstract
Conversational AI has become an integral part of everyday life.Historical evidence shows that sociocultural change can be reflected in mental health, particularly in the incorporation of new technologies, suggesting that conversational AI represents a new context through which such changes may influence the manifestation and content of psychotic symptoms.This narrative review explores the possible relationship between conversational AI use and the development or worsening of psychotic symptoms, bringing together the limited clinical evidence, possible neurocognitive mechanisms, relevant vulnerability factors, and the potential role of conversational AI in therapeutic and pathological contexts.A non-systematic review of the literature was conducted using PubMed and Google Scholar, with relevant clinical case reports, original research articles, reviews, and conceptual articles identified through keyword searches and reference list screening.Conversational AI use has been associated with psychotic-like experiences in cross-sectional survey data and has been implicated in the onset or relapse of psychotic symptoms in an isolated case report.Several theoretical neurocognitive mechanisms have also been proposed by which conversational AI could contribute to the emergence or worsening of psychotic symptoms in vulnerable individuals.These mechanisms are largely derived from existing psychosis models and remain hypothesis-generating rather than empirically demonstrated, requiring further validation through longitudinal and experimental studies.At the same time, in carefully defined clinical contexts, conversational AI may also have supportive uses in assessment, monitoring, or structured therapeutic support.Given the current context, the use of conversational AI should be approached with caution; adherence to ethical standards and careful consideration of both its benefits and associated psychological risks are vital.
Citations in the library
Cites 7
- Psychosis as a state of aberrant salience: a framework linking biology, phenomenology, and pharmacology in schizophrenia. American Journal of Psychiatry 2003
- The Predictive Coding Account of Psychosis. Biological Psychiatry May 25, 2018
- 20 Years of Aberrant Salience in Psychosis: What Have We Learned? American Journal of Psychiatry March 26, 2025
- 'The algorithm is hacked': analysis of technology delusions in a modern-day cohort. The British Journal of Psychiatry November 3, 2025
- Delusional Experiences Emerging From AI Chatbot Interactions or "AI Psychosis". JMIR Mental Health December 3, 2025
and 2 more in the library