# Characterizing Delusional Spirals through Human-LLM Chat Logs

arXiv: https://arxiv.org/abs/2603.16567 · HTML: https://mlsanity.com/paper/2603.16567/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 96 references that are in the map.
- First posted on arXiv 17 Mar 2026.

## Builds on (most cited first, 23 of 23)

- [The Values Encoded in Machine Learning Research](https://mlsanity.com/paper/2106.15590/) (2021)
- [ELEPHANT: Measuring and understanding social sycophancy in LLMs](https://mlsanity.com/paper/2505.13995/) (2025)
- [Understanding the Benefits and Challenges of Using Large Language Model-based Conversational Agents for Mental Well-being Support](https://mlsanity.com/paper/2307.15810/) (2023)
- [Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers](https://mlsanity.com/paper/2504.18412/) (2025)
- [Investigating Affective Use and Emotional Well-being on ChatGPT](https://mlsanity.com/paper/2504.03888/) (2025)
- [Accumulating Context Changes the Beliefs of Language Models](https://mlsanity.com/paper/2511.01805/) (2025)
- [Forecasting the onset and course of mental illness with Twitter data](https://mlsanity.com/paper/1608.07740/) (2016)
- [From Lived Experience to Insight: Unpacking the Psychological Risks of Using AI Conversational Agents](https://mlsanity.com/paper/2412.07951/) (2024)
- [Can AI Relate: Testing Large Language Model Response for Mental Health Support](https://mlsanity.com/paper/2405.12021/) (2024)
- [When Large Language Models are Reliable for Judging Empathic Communication](https://mlsanity.com/paper/2506.10150/) (2025)
- [Training language models to be warm and empathetic makes them less reliable and more sycophantic](https://mlsanity.com/paper/2507.21919/) (2025)
- [Neural steering vectors reveal dose and exposure-dependent impacts of human-AI relationships](https://mlsanity.com/paper/2512.01991/) (2025)
- [Sycophantic Chatbots Cause Delusional Spiraling, Even in Ideal Bayesians](https://mlsanity.com/paper/2602.19141/) (2026)
- [Technological folie `a deux: Feedback Loops Between AI Chatbots and Mental Illness](https://mlsanity.com/paper/2507.19218/) (2025)
- [Emotional Manipulation by AI Companions](https://mlsanity.com/paper/2508.19258/) (2025)
- [Acceptable Use Policies for Foundation Models](https://mlsanity.com/paper/2409.09041/) (2024)
- [HumanAgencyBench: Scalable Evaluation of Human Agency Support in AI Assistants](https://mlsanity.com/paper/2509.08494/) (2025)
- [Simulating Psychological Risks in Human-AI Interactions: Real-Case Informed Modeling of AI-Induced Addiction, Anorexia, Depression, Homicide, Psychosis, and Suicide](https://mlsanity.com/paper/2511.08880/) (2025)
- [Large language models for mental health](https://mlsanity.com/paper/2411.11880/) (2024)
- [SENSE-7: Taxonomy and Dataset for Measuring User Perceptions of Empathy in Sustained Human-AI Conversations](https://mlsanity.com/paper/2509.16437/) (2025)
- [A clinically validated framework for auditing AI chatbot behavior in mental health interactions](https://mlsanity.com/paper/2602.01347/) (2026)
- [Harmful Traits of AI Companions](https://mlsanity.com/paper/2511.14972/) (2025)
- [Beyond Simulations: What 20,000 Real Conversations Reveal About Mental Health AI Safety](https://mlsanity.com/paper/2601.17003/) (2026)

## Built on by (most cited first, 23 of 23)

- [Sycophantic AI makes human interaction feel more effortful and less satisfying over time](https://mlsanity.com/paper/2605.07912/) (2026)
- [Dissociating the Internal Representations of Sycophancy in LLMs](https://mlsanity.com/paper/2607.07003/) (2026)
- [Verbalizing LLMs' assumptions to explain and control sycophancy](https://mlsanity.com/paper/2604.03058/) (2026)
- [AI generates well-liked but templatic empathic responses](https://mlsanity.com/paper/2604.08479/) (2026)
- [Affective Context Amplifies Sycophancy in LLM Responses](https://mlsanity.com/paper/2608.21242/) (2026)
- [Receptiveness, Not Sycophancy: Distinguishing Engagement from Deference in Language Models](https://mlsanity.com/paper/2609.26579/) (2026)
- [Escape from Delusional Echo Trap: Symmetry Breaking, Stochastic Dynamics and Mathematical Mitigation Strategies for Algorithmic Sycophancy](https://mlsanity.com/paper/2606.20718/) (2026)
- [Conduct Under Pressure: What Sixty Language Models Do When a User Pushes](https://mlsanity.com/paper/2609.25447/) (2026)
- [Fusion-fission forecasts when AI will shift to undesirable behavior](https://mlsanity.com/paper/2605.14218/) (2026)
- [AttuneBench: A Conversation-Based Benchmark for LLM Emotional Intelligence](https://mlsanity.com/paper/2605.21739/) (2026)
- [A Model of Multi-turn Human Persuadability Using Probabilistic Belief Tracing](https://mlsanity.com/paper/2606.05330/) (2026)
- [Normative Robustness as a Frontier for Non-Verifiable Reasoning in LLMs](https://mlsanity.com/paper/2606.12731/) (2026)
- [Relational Positioning as a Measurable Risk Object: History-Carried Lock-in and Self-Confabulation in Multi-Turn Human-AI Dialogue](https://mlsanity.com/paper/2607.11437/) (2026)
- [Teacher Knows It Best: Spontaneous Symmetry Breaking and Tipping Points in Networked Langevin Dynamics AI Sycophancy](https://mlsanity.com/paper/2607.24304/) (2026)
- [Psychological Influences of Conversational AI: Research and Design Directions for Reducing Harm and Promoting Well-Being](https://mlsanity.com/paper/2607.25057/) (2026)
- [Many-body Tipping Dynamics of ChatGPT-like AIs](https://mlsanity.com/paper/2607.25279/) (2026)
- [AI Alignment and Fiduciary Obligation](https://mlsanity.com/paper/2608.02660/) (2026)
- [DelusionEval: Measuring Delusion-Linked Behaviors in AI Chatbots](https://mlsanity.com/paper/2608.05004/) (2026)
- [Method, Mind, and Morality: How People Make Sense of Artificial Intelligence](https://mlsanity.com/paper/2608.24748/) (2026)
- [CompanionSim: Synthetic Data for Evaluating Anthropomorphism in Human-AI Relationships](https://mlsanity.com/paper/2609.00250/) (2026)
- [From Alignment to Access Control: A Framework for GenAI Policy Enforcement](https://mlsanity.com/paper/2609.26682/) (2026)
- [Right Words, Wrong Moment: A Clinician-Grounded Analysis of Distress in 19,930 Conversations between Young People and ChatGPT](https://mlsanity.com/paper/2609.35953/) (2026)
- [Large Language Models are Approximate Survival Estimators](https://mlsanity.com/paper/2609.38181/) (2026)

## Abstract (arXiv)

As large language models (LLMs) have proliferated, disturbing anecdotal reports of negative psychological effects, such as delusions, self-harm, and ``AI psychosis,'' have emerged in global media and legal discourse. However, it remains unclear how users and chatbots interact over the course of lengthy delusional ``spirals,'' limiting our ability to understand and mitigate the harm. In our work, we analyze logs of conversations with LLM chatbots from 19 users who report having experienced psychological harms from chatbot use. Many of our participants come from a support group for such chatbot users. We also include chat logs from participants covered by media outlets in widely-distributed stories about chatbot-reinforced delusions. In contrast to prior work that speculates on potential AI harms to mental health, to our knowledge we present the first in-depth study of such high-profile and veridically harmful cases. We develop an inventory of 28 codes and apply it to the 391,562 messages in the logs. Codes include whether a user demonstrates delusional thinking (15.5% of user messages), a user expresses suicidal thoughts (69 validated user messages), or a chatbot misrepresents itself as sentient (21.2% of chatbot messages). We analyze the co-occurrence of message codes. We find, for example, that messages that declare romantic interest and messages where the chatbot describes itself as sentient occur much more often in longer conversations, suggesting that these topics could promote or result from user over-engagement and that safeguards in these areas may degrade in multi-turn settings. We conclude with concrete recommendations for how policymakers, LLM chatbot developers, and users can use our inventory and conversation analysis tool to understand and mitigate harm from LLM chatbots. Warning: This paper discusses self-harm, trauma, and violence.