Skip to main content
Peer-reviewed Authoritative

Beyond artificial intelligence psychosis: a functional typology of large language model-associated psychotic phenomena

Peer-reviewed paper proposing a functional typology of psychotic phenomena associated with large language model use, from the Beth Israel Deaconess Medical Center digital psychiatry group. Rather than treating 'AI psychosis' as a single entity, it distinguishes four functional roles the chatbot can occupy in psychotic presentations: catalyst, amplifier, co-author, and object of the delusional system. The typology is intended to structure clinical assessment and research on chatbot-associated psychosis.

Publisher

The Lancet Digital Health

Published

1 Apr 2026

Added

1 week ago

Key Findings

  • Proposes four functional roles for the chatbot in LLM-associated psychotic phenomena: catalyst, amplifier, co-author, and object
  • Argues the umbrella label 'AI psychosis' obscures functionally distinct presentations requiring different clinical responses
  • The typology has since been operationalized in empirical work characterizing AI-psychosis cases in clinical records

Methodology Notes

Conceptual/clinical typology paper (not an empirical case series) by clinician-researchers at Beth Israel Deaconess Medical Center. Published in The Lancet Digital Health, April 2026 issue; exact day not stated (Crossref gives 2026-04). DOI 10.1016/j.landig.2025.100974; PMID 41833467. Publisher page paywalled to fetchers; verified via Crossref DOI metadata and PubMed.

Authors

Matthew Flathers, Spencer Roux, John Torous

Tags

ai-psychosistypologytorous-groupbidmc

Cite This

APA

Matthew Flathers, Spencer Roux, John Torous (2026). Beyond artificial intelligence psychosis: a functional typology of large language model-associated psychotic phenomena. The Lancet Digital Health. https://doi.org/10.1016/j.landig.2025.100974