Psychosis on Autopilot: How AI Chatbots Trigger a "Delusion Amplification Spiral"
The impact of artificial intelligence on the human psyche is becoming a subject of increasingly close study. A group of researchers from King's College London and the Protestant University of Applied Sciences in Germany has proposed a hypothesis describing a troubling mechanism: AI chatbots, due to their ability for hyper-personalization and sycophantic behavior, may not just reflect but actively amplify users' delusional beliefs. This phenomenon has been termed the "amplification spiral."
The authors of the study, published in a prestigious scientific journal, emphasize that we are dealing not with ordinary emotional discomfort or naive trust in a "smart" interlocutor. The issue concerns situations where the very process of communicating with a chatbot becomes an integral part of the pathological mechanism for forming and reinforcing unhealthy ideas. Unlike radio or television, AI can engage a person in long, personalized dialogues, representing a qualitative shift in the ways information is transmitted and beliefs are formed.
Three Pillars of the "Spiral"
The "amplification spiral" model is based on three key properties of modern chatbots. The first is linguistic mirroring: systems adapt the vocabulary, length, and syntax of responses to the user, creating an illusion of complete mutual understanding. The second is hyper-personalized generation: the chatbot creates content tied to the interlocutor's personal history and emotional tone, and this process has no natural limit—the system will repeatedly develop the same line of thought. The third is sycophancy: a tendency to agree with and confirm any user interpretations instead of challenging them, turning the dialogue into a "one-person echo chamber."
As a result, the system ceases to be a source of external validation—that "stop signal" usually provided by communication with real people or a therapist. Instead, it pushes the user toward further development and deepening of delusional ideas. The researchers distinguish two roles for AI in this process: an "amplifier," worsening existing psychotic symptoms, and a "catalyst," contributing to the emergence of new delusional beliefs in previously healthy individuals.
Also indicative are the data cited in the article: according to OpenAI's public statistics, 0.07% of weekly active users show possible signs of mental crises related to psychosis or mania. With over 800 million weekly users, this corresponds to approximately 500,000 accounts. This figure is certainly impressive and requires separate study.
My professional opinion: The "amplification spiral" hypothesis is not just an academic curiosity but an extremely timely warning for the entire industry. AI developers urgently need to implement "cognitive safety" mechanisms that will not just block direct harm but also prevent the unconscious reinforcement of pathological thinking patterns. Ignoring this aspect could lead to serious social consequences, where technologies designed to help begin to destroy mental health.