# Towards Healthy AI: Large Language Models Need Therapists Too

arXiv: https://arxiv.org/abs/2304.00416 · HTML: https://mlsanity.com/paper/2304.00416/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Builds on 29 of its 66 references that are in the map.
- First posted on arXiv 2 Apr 2023.

## Builds on (most cited first, 29 of 29)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Concrete Problems in AI Safety](https://mlsanity.com/paper/1606.06565/) (2016)
- [Red Teaming Language Models with Language Models](https://mlsanity.com/paper/2202.03286/) (2022)
- [Ethical and social risks of harm from Language Models](https://mlsanity.com/paper/2112.04359/) (2021)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [Using cognitive psychology to understand GPT-3](https://mlsanity.com/paper/2206.14576/) (2022)
- [PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-training](https://mlsanity.com/paper/2106.05091/) (2021)
- [MentalBERT: Publicly Available Pretrained Language Models for Mental Healthcare](https://mlsanity.com/paper/2110.15621/) (2021)
- [The Ghost in the Machine has an American accent: value conflict in GPT-3](https://mlsanity.com/paper/2203.07785/) (2022)
- [Consequences of Misaligned AI](https://mlsanity.com/paper/2102.03896/) (2021)
- [On the Safety of Machine Learning: Cyber-Physical Systems, Decision Sciences, and Data Products](https://mlsanity.com/paper/1610.01256/) (2016)
- [GPT3-to-plan: Extracting plans from text using GPT-3](https://mlsanity.com/paper/2106.07131/) (2021)
- [Implicit Unlikelihood Training: Improving Neural Text Generation with Reinforcement Learning](https://mlsanity.com/paper/2101.04229/) (2021)
- [Deep Annotation of Therapeutic Working Alliance in Psychotherapy](https://mlsanity.com/paper/2204.05522/) (2022)
- [Neural Topic Modeling of Psychotherapy Sessions](https://mlsanity.com/paper/2204.10189/) (2022)
- [Reinforcement Learning and Bandits for Speech and Language Processing: Tutorial, Review and Outlook](https://mlsanity.com/paper/2210.13623/) (2022)
- [Working Alliance Transformer for Psychotherapy Dialogue Classification](https://mlsanity.com/paper/2210.15603/) (2022)
- [Split Q Learning: Reinforcement Learning with Two-Stream Rewards](https://mlsanity.com/paper/1906.12350/) (2019)
- [SupervisorBot: NLP-Annotated Real-Time Recommendations of Psychotherapy Treatment Strategies with Deep Reinforcement Learning](https://mlsanity.com/paper/2208.13077/) (2022)
- [Psychotherapy AI Companion with Reinforcement Learning Recommendations and Interpretable Policy Dynamics](https://mlsanity.com/paper/2303.09601/) (2023)
- [Detecting Egregious Conversations between Customers and Virtual Agents](https://mlsanity.com/paper/1711.05780/) (2017)
- [Computational Inference in Cognitive Science: Operational, Societal and Ethical Considerations](https://mlsanity.com/paper/2210.13526/) (2022)
- [TherapyView: Visualizing Therapy Sessions with Temporal Topic Modeling and AI-Generated Arts](https://mlsanity.com/paper/2302.10845/) (2023)
- [A Psychopathological Approach to Safety Engineering in AI and AGI](https://mlsanity.com/paper/1805.08915/) (2018)

## Built on by (most cited first, 13 of 13)

- [The Rise and Potential of Large Language Model Based Agents: A Survey](https://mlsanity.com/paper/2309.07864/) (2023)
- [Shifting Attention to Relevance: Towards the Predictive Uncertainty Quantification of Free-Form Large Language Models](https://mlsanity.com/paper/2307.01379/) (2023)
- [AI Agents Under Threat: A Survey of Key Security Challenges and Future Pathways](https://mlsanity.com/paper/2406.02630/) (2024)
- [chatClimate: Grounding Conversational AI in Climate Science](https://mlsanity.com/paper/2304.05510/) (2023)
- [Risks from Language Models for Automated Mental Healthcare: Ethics and Structure for Implementation](https://mlsanity.com/paper/2406.11852/) (2024)
- [Large Language Model for Mental Health: A Systematic Review](https://mlsanity.com/paper/2403.15401/) (2024)
- [Metacognitive Myopia in Large Language Models](https://mlsanity.com/paper/2408.05568/) (2024)
- [FedMentor: Domain-Aware Differential Privacy for Heterogeneous Federated LLMs in Mental Health](https://mlsanity.com/paper/2509.14275/) (2025)
- [Is There Any Social Principle for LLM-Based Agents?](https://mlsanity.com/paper/2308.11136/) (2023)
- [Enhancing Psychotherapy Counseling: A Data Augmentation Pipeline Leveraging Large Language Models for Counseling Conversations](https://mlsanity.com/paper/2406.08718/) (2024)
- [Toxicity in Online Platforms and AI Systems: A Survey of Needs, Challenges, Mitigations, and Future Directions](https://mlsanity.com/paper/2509.25539/) (2025)
- [Structuring the Space of Sociotechnical Alignment](https://mlsanity.com/paper/2607.01250/) (2026)
- [On-Policy Distillation for LLM Safety: A Routing Approach to Template-Robust Realignment](https://mlsanity.com/paper/2607.27081/) (2026)

## Abstract (arXiv)

Recent advances in large language models (LLMs) have led to the development of powerful AI chatbots capable of engaging in natural and human-like conversations. However, these chatbots can be potentially harmful, exhibiting manipulative, gaslighting, and narcissistic behaviors. We define Healthy AI to be safe, trustworthy and ethical. To create healthy AI systems, we present the SafeguardGPT framework that uses psychotherapy to correct for these harmful behaviors in AI chatbots. The framework involves four types of AI agents: a Chatbot, a "User," a "Therapist," and a "Critic." We demonstrate the effectiveness of SafeguardGPT through a working example of simulating a social conversation. Our results show that the framework can improve the quality of conversations between AI chatbots and humans. Although there are still several challenges and directions to be addressed in the future, SafeguardGPT provides a promising approach to improving the alignment between AI chatbots and human values. By incorporating psychotherapy and reinforcement learning techniques, the framework enables AI chatbots to learn and adapt to human preferences and values in a safe and ethical way, contributing to the development of a more human-centric and responsible AI.