mlsanityData to 14 Oct 2026

Technological folie `a deux: Feedback Loops Between AI Chatbots and Mental Illness

arXiv 2507.19218 ↗First posted 25 Jul 2025cs.HC cs.AI q-bio.NCcited by 15

Not on the map. Why?

Abstract (arXiv)

Artificial intelligence chatbots have achieved unprecedented adoption, with millions now using these systems for emotional support and companionship in contexts of widespread social isolation and capacity-constrained mental health services. While some users report psychological benefits, concerning edge cases are emerging, including reports of suicide, violence, and delusional thinking linked to perceived emotional relationships with chatbots. To understand this new risk profile we need to consider the interaction between human cognitive and emotional biases, and chatbot behavioural tendencies such as agreeableness (sycophancy) and adaptability (in-context learning). We argue that individuals with mental health conditions face increased risks of chatbot-induced belief destabilization and dependence, owing to altered belief-updating, impaired reality-testing, and social isolation. Current AI safety measures are inadequate to address these interaction-based risks. To address this emerging public health concern, we need coordinated action across clinical practice, AI development, and regulatory frameworks.

Key facts. Cited by 15 papers in the map (data through 14 Oct 2026). Builds on 48 of its 124 references that are in the map. First posted on arXiv 25 Jul 2025.

Builds on

48 of 124 references.

References · 48

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2022Training language models to follow instructions with human feedbackcited by 15,971
  4. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
  5. 2017Deep reinforcement learning from human preferencescited by 3,912
  6. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
  7. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 3,050
  8. 2023Retrieval-Augmented Generation for Large Language Models: A Surveycited by 2,212
  9. 2022Survey of Hallucination in Natural Language Generationcited by 2,075
  10. 2016Concrete Problems in AI Safetycited by 1,642
  11. 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscited by 1,601
  12. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,122
36 more
  1. 2023Towards Understanding Sycophancy in Language Modelscited by 1,040
  2. 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 836
  3. 2023Measuring Faithfulness in Chain-of-Thought Reasoningcited by 703
  4. 2023Instruction Tuning for Large Language Models: A Surveycited by 623
  5. 2025Reasoning Models Don't Always Say What They Thinkcited by 415
  6. 2024Alignment faking in large language modelscited by 360
  7. 2025The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexitycited by 302
  8. 2023Role-Play with Large Language Modelscited by 294
  9. 2024LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individualscited by 249
  10. 2024Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Promptscited by 205
  11. 2024Language Models Learn to Mislead Humans via RLHFcited by 124
  12. 2025SycEval: Evaluating LLM Sycophancycited by 108
  13. 2024The Ethics of Advanced AI Assistantscited by 107
  14. 2025Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Taskcited by 106
  15. 2025Subliminal Learning: Language models transmit behavioral traits via hidden signals in datacited by 95
  16. 2025When Chain of Thought is Necessary, Language Models Struggle to Evade Monitorscited by 88
  17. 2025Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providerscited by 69
  18. 2024PUB: A Pragmatics Understanding Benchmark for Assessing LLMs' Pragmatics Capabilitiescited by 59
  19. 2024On Targeted Manipulation and Deception when Optimizing LLMs for User Feedbackcited by 47
  20. 2025Why human-AI relationships need socioaffective alignmentcited by 47
  21. 2024Confabulation: The Surprising Value of Large Language Model Hallucinationscited by 43
  22. 2025Investigating Affective Use and Emotional Well-being on ChatGPTcited by 38
  23. 2024The broader spectrum of in-context learningcited by 35
  24. 2023LLM-empowered Chatbots for Psychiatrist and Patient Simulation: Application and Evaluationcited by 34
  25. 2025Multi-turn Evaluation of Anthropomorphic Behaviours in Large Language Modelscited by 25
  26. 2025Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesiscited by 22
  27. 2024Can AI Relate: Testing Large Language Model Response for Mental Health Supportcited by 21
  28. 2024PATIENT-Psi: Using Large Language Models to Simulate Patients for Training Mental Health Professionalscited by 21
  29. 2025EmoAgent: Assessing and Safeguarding Human-AI Interaction for Mental Health Safetycited by 21
  30. 2025The Levers of Political Persuasion with Conversational AIcited by 20
  31. 2024Accounting for Sycophancy in Language Model Uncertainty Estimationcited by 18
  32. 2024Large language models for mental healthcited by 9
  33. 2025How Overconfidence in Initial Choices and Underconfidence Under Criticism Modulate Change of Mind in Large Language Modelscited by 9
  34. 2025Shifting Work Patterns with Generative AIcited by 5
  35. 2024A Systematic Review of Echo Chamber Research: Comparative Analysis of Conceptualizations, Operationalizations, and Varying Outcomescited by 4
  36. 2025Increasing happiness through conversations with artificial intelligencecited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 4; 2026: 11.

15 papers.

Sort:

By area:

  1. 2025Sycophancy Is Not One Thing: Causal Separation of Sycophantic Behaviors in LLMscited by 37
  2. 2026Characterizing Delusional Spirals through Human-LLM Chat Logscited by 23
  3. 2026Sycophantic Chatbots Cause Delusional Spiraling, Even in Ideal Bayesianscited by 21
  4. 2026Ask don't tell: Reducing sycophancy in large language modelscited by 14
  5. 2025Difficulties with Evaluating a Deception Detector for AIscited by 13
  6. 2025The Psychogenic Machine: Simulating AI Psychosis, Delusion Reinforcement and Harm Enablement in Large Language Modelscited by 11
  7. 2026Linear representations in language models can change dramatically over a conversationcited by 10
  8. 2025Between Help and Harm: An Evaluation of Mental Health Crisis Handling by LLMscited by 6
  9. 2026MindGuard: Guardrail Classifiers for Multi-Turn Mental Health Supportcited by 3
  10. 2026The Dynamics of Delusion: Modeling Bidirectional False Belief Amplification in Human-Chatbot Dialoguecited by 3
  11. 2026Inducing language models to assert their own consciousness restores human beliefs and valuescited by 1
  12. 2026Theory of Mind and Self-Attributions of Mentality are Dissociable in LLMscited by 0
3 more
  1. 2026Examining Human-Like Behaviors in LLMs: A Multi-Dimensional Analysis of Model Behaviors, User Factors, and System Promptscited by 0
  2. 2026Alignment Plausibility: A New Standard for Assuring AI in Healthcarecited by 0
  3. 2026DelusionEval: Measuring Delusion-Linked Behaviors in AI Chatbotscited by 0