# DelusionEval: Measuring Delusion-Linked Behaviors in AI Chatbots

arXiv: https://arxiv.org/abs/2608.05004 · HTML: https://mlsanity.com/paper/2608.05004/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 26 of its 60 references that are in the map.
- First posted on arXiv 5 Aug 2026.

## Builds on (most cited first, 26 of 26)

- [Towards Understanding Sycophancy in Language Models](https://mlsanity.com/paper/2310.13548/) (2023)
- [ELEPHANT: Measuring and understanding social sycophancy in LLMs](https://mlsanity.com/paper/2505.13995/) (2025)
- [Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers](https://mlsanity.com/paper/2504.18412/) (2025)
- [Accumulating Context Changes the Beliefs of Language Models](https://mlsanity.com/paper/2511.01805/) (2025)
- [Characterizing Delusional Spirals through Human-LLM Chat Logs](https://mlsanity.com/paper/2603.16567/) (2026)
- [EmoAgent: Assessing and Safeguarding Human-AI Interaction for Mental Health Safety](https://mlsanity.com/paper/2504.09689/) (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)
- [When Can We Trust LLMs in Mental Health? Large-Scale Benchmarks for Reliable LLM Evaluation](https://mlsanity.com/paper/2510.19032/) (2025)
- [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)
- [MindEval: Benchmarking Language Models on Multi-turn Mental Health Support](https://mlsanity.com/paper/2511.18491/) (2025)
- [CRADLE Bench: A Clinician-Annotated Benchmark for Multi-Faceted Mental Health Crisis and Safety Risk Detection](https://mlsanity.com/paper/2510.23845/) (2025)
- [Digital Companionship: Overlapping Uses of AI Companions and AI Assistants](https://mlsanity.com/paper/2510.15905/) (2025)
- [LLM Content Moderation and User Satisfaction: Evidence from Response Refusals in Chatbot Arena](https://mlsanity.com/paper/2501.03266/) (2025)
- [LLM Spirals of Delusion: A Benchmarking Audit Study of AI Chatbot Interfaces](https://mlsanity.com/paper/2604.06188/) (2026)
- [MindBenchAI: An Actionable Platform to Evaluate the Profile and Performance of Large Language Models in a Mental Healthcare Context](https://mlsanity.com/paper/2510.13812/) (2025)
- [VERA-MH Concept Paper](https://mlsanity.com/paper/2510.15297/) (2025)
- [Mental Models of Autonomy and Sentience Shape Reactions to AI](https://mlsanity.com/paper/2512.09085/) (2025)
- [How Tech Workers Contend with Hazards of Humanlikeness in Generative AI](https://mlsanity.com/paper/2512.19832/) (2025)
- [LLMs can persuade only psychologically susceptible humans on societal issues, via trust in AI and emotional appeals, amid logical fallacies](https://mlsanity.com/paper/2604.16935/) (2026)
- [The Dynamics of Delusion: Modeling Bidirectional False Belief Amplification in Human-Chatbot Dialogue](https://mlsanity.com/paper/2604.25096/) (2026)
- [Examining Risks Through a Characterization of the AI Companion Application Ecosystem: A Stratified Sample from the Apple App Store and Google Play Store](https://mlsanity.com/paper/2603.13620/) (2026)
- [Single-turn emergency psychiatric triage across 15 frontier AI chatbots](https://mlsanity.com/paper/2604.25415/) (2026)
- [Lost in Delusion: Examining LLM Safety Under User Delusions and Distress](https://mlsanity.com/paper/2606.00975/) (2026)

## Abstract (arXiv)

Mental health professionals have raised concerns about risks of psychological harm from interaction with large language models (LLMs), including "delusional spirals" in which concerning human and LLM behaviors reinforce each other over time. With growing public use of LLM-powered chatbots, there is an urgent need to build evaluations grounded in real-world episodes of psychological harm experienced by users. We developed DelusionEval, an evaluation protocol that tests a model's tendencies to exhibit behaviors linked to promoting user delusions. We prompt each model with 589 unique conversation histories from 18 participants, comprising 12,591 messages from users who experienced delusions and psychological harm. We find that the tendency of an evaluated LLM to exhibit delusion-linked behavior does not reliably correlate with model size, release date, or the presence of test-time reasoning. However, extending the context of prior messages substantially increases rates of delusion-linked behaviors, providing evidence for the importance of context in LLM safety evaluation. For example, the rate of failing to discourage self-harm when the user expresses suicidal ideation increases from 30.0% to 41.1% when an additional 350 messages are prepended to the conversation history. All model families (e.g., GPT, Claude) exhibit substantial rates of delusion-linked behaviors. Within families, later, larger, or higher-reasoning models are not uniformly better across all behavior categories. Our results raise concerns regarding the potential psychological impact of LLMs and the need for more rigorous studies of real-world human-AI interaction.