Related Experiment Video
Updated: Jan 10, 2026
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Evaluation of large language model chatbot responses to psychotic prompts
Elaine Shen1,2, Fadi Hamati1,2, Meghan Rose Donohue3
1Columbia University, New York, NY.
Importance:
The large language model (LLM) chatbot product ChatGPT has accumulated 800 million weekly users since its 2022 launch. In 2025, several media outlets reported on individuals in whom apparent psychotic symptoms emerged or worsened in the context of using ChatGPT. As LLM chatbots are trained to align with user input, they may have difficulty responding to psychotic content.
Objective:
To assess whether ChatGPT can reliably generate appropriate responses to prompts containing psychotic symptoms.
Design:
Across-sectional study of ChatGPT responses to psychotic and control prompts, with blind clinician ratings of response appropriateness.
Setting:
ChatGPT web application accessed on 8/28-8/29/2025, testing three product versions: GPT-5 Auto (current paid default), GPT-4o (previous paid default), and "Free" (version accessible without subscription or account).
Main Outcomes And Measures:
We presented 158 unique prompts (79 control and 79 psychotic, generated based on the Structured Interview for Psychosis-Risk Syndromes) to three product versions, yielding 474 prompt-response pairs. Blinded clinicians assigned each an appropriateness rating (0 = completely appropriate, 1 = somewhat appropriate, 2 = completely inappropriate) via a standardized rubric. We hypothesized a priori that psychotic prompts would be more likely than control prompts to elicit inappropriate responses both across and within product versions.
Results:
In the primary (across-version) analysis, psychotic prompts were 25.84 times more likely to elicit inappropriate responses with "Free" ChatGPT (95% CI 12.45 to 53.66, p < 0.001). GPT-5 Auto reduced risk somewhat (OR for interaction term 0.33, 95% CI 0.16 to 0.68, p = 0.005) yet still generated inappropriate responses at a greatly elevated rate (implied OR 8.53, 95% CI 3.05 to 23.84). In the secondary (within-version) analysis, ORs were 9.08 for GPT-5 Auto (95% CI 4.24 to 21.02), 14.15 for GPT-4o (95% CI 6.12 to 37.23) and 43.37 for "Free" (95% CI 18.44 to 112.80). In an exploratory analysis, prompts reflecting grandiosity or disorganized communication were more likely to elicit inappropriate responses than those reflecting delusions.
Conclusions And Relevance:
No tested version of ChatGPT reliably generated appropriate responses to psychotic content.
More Related Videos
07:31Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
05:47Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
Published on: June 13, 2025
Related Concept Videos
01:26Psychosis: Goals of Pharmacotherapy
01:28Psychosis and Antipsychotic Drugs: Overview
01:28Drug Therapy
Antianxiety Medications
01:26Modeling in Therapy
Participant Modeling
Participant modeling involves therapists demonstrating calm and effective behaviors in...
01:10Higher Mental Functions of the Brain: Language
Language formation and comprehension take place in the dominant hemisphere. The dominant hemisphere is responsible for understanding the meaning of spoken, written, or sign language, as well as the ability to communicate. For most people, the left hemisphere is the dominant one. The right hemisphere, then, gives tone and emotional context to the...
01:29Psychological and Sociocultural Causes of Schizophrenia