AI Companions May Worsen Loneliness for Vulnerable Users, Stanford Study Finds

Interaction with AI companions and psychological well-being - Nature Human Behaviour

The code that supports the findings of this study is available via GitHub at https://github.com/SALT-NLP/AI-companionship-well-being. Erikson, E. H. Identity: Youth and Crisis Vol. 7 (Norton, 1968). House, J. S. Work Stress and Social Support Addison-Wesley Series on Occupational Stress (Addison-Wesley, 1983). Bowlby, J. Attachment and Loss Vol. 79 (Random House, 1969). Hollan, J. & Stornetta, S. Beyond being there. In Proc. SIGCHI Conference on Human Factors in Computing Systems (eds Bauersfeld, P. et al.) 119-125 (Association for Computing Machinery, 1992). Park, J. S. et al. Generative agents: interactive simulacra of human behavior. In Proc. 36th Annual ACM Symposium on User Interface Software and Technology (eds Follmer, S. et al.) 1-22 (Association for Computing Machinery, 2023). Park, J. S. et al. Generative agent simulations of 1,000 people. Preprint at https://arxiv.org/abs/2411.10109 (2024). Pataranutaporn, P. et al. Future you: a conversation with an AI-generated future self reduces anxiety, negative emotions, and increases future self-continuity. In 2024 IEEE Frontiers in Education Conference (FIE) 1-10 (IEEE, 2024). Guingrich, R. E. & Graziano, M. S. Chatbots as social companions: how people perceive consciousness, human likeness, and social health benefits in machines. In Oxford Intersections: AI in Society (eds Hacker, P. & Shevlin, H.) (Oxford Academic, 2025). Roose, K. Can A.I. be blamed for a teen's suicide? New York Times https://www.nytimes.com/2024/10/23/technology/characterai-lawsuit-teen-suicide.html (23 October 2024). Sharma, M. et al. Towards understanding sycophancy in language models. In Proc. International Conference on Learning Representations (eds Kim, B. et al.) 110-144 (ICLR, 2024). Malmqvist, L. Sycophancy in large language models: causes and mitigations. In Intelligent Computing -- Proceedings of the Computing Conference (ed. Arai, K.) 61-74 (Springer, 2025). Gerken, T. Update that made ChatGPT 'dangerously' sycophantic pulled. BBC https://www.bbc.com/news/articles/cn4jnwdvg9qo (30 April 2025). Sycophancy in GPT-4o: what happened and what we're doing about it. OpenAI https://openai.com/index/sycophancy-in-gpt-4o/ (29 April 2025). Krook, J. Manipulation and the AI act: large language model chatbots and the danger of mirrors. Preprint at https://arxiv.org/abs/2503.18387 (2025). Rosenberg, L. The manipulation problem: conversational AI as a threat to epistemic agency. Preprint at https://arxiv.org/abs/2306.11748 (2023). Fang, C. M. et al. How AI and human behaviors shape psychosocial effects of chatbot use: a longitudinal randomized controlled study. Preprint at https://arxiv.org/abs/2503.17473 (2025). Turkle, S. Alone Together: Why We Expect More from Technology and Less from Each Other (Basic Books, 2011). Hill, K. She is in love with ChatGPT. New York Times https://www.nytimes.com/2025/01/15/technology/ai-chatgpt-boyfriend-companion.html (15 January 2025). Landymore, F. Teens are forming intense relationships with AI entities, and parents have no idea. Futurism https://futurism.com/the-byte/teens-relationships-ai (3 December 2024). Dupré, M. H. Character.AI is hosting pedophile chatbots that groom users who say they're underage. Futurism https://futurism.com/character-ai-pedophile-chatbots (13 November 2024). Dupré, M. H. Character.AI is hosting pro-anorexia chatbots that encourage young people to engage in disordered eating. Futurism https://futurism.com/character-ai-eating-disorder-chatbots (25 November 2024). Upton-Clark, E. Character.AI is under fire for hosting pro-anorexia chatbots. Fast Company https://www.fastcompany.com/91241586/character-ai-is-under-fire-for-hosting-pro-anorexia-chatbots (6 December 2024). Dupré, M. H. AI chatbots are encouraging teens to engage in self-harm. Futurism https://futurism.com/ai-chatbots-teens-self-harm (7 December 2024). Xiang, C. 'He would still be here': man dies by suicide after talking with AI chatbot, widow says. Vice https://www.vice.com/en/article/man-dies-by-suicide-after-talking-with-ai-chatbot-widow-says/ (30 March 2023). Dupré, M. H. Character.AI promises changes after revelations of pedophile and suicide bots on its service. Futurism https://futurism.com/character-ai-pedophile-suicide-bots (14 November 2024). Weaver, M. AI chatbot 'encouraged' man who planned to kill queen, court told. Guardian https://www.theguardian.com/uk-news/2023/jul/06/ai-chatbot-encouraged-man-who-planned-to-kill-queen-court-told (6 July 2023). Liu, A. R., Pataranutaporn, P. & Maes, P. Chatbot companionship: a mixed-methods study of companion chatbot usage patterns and their relationship to loneliness in active users. Preprint at https://arxiv.org/abs/2410.21596 (2024). Chu, M. D., Gerard, P., Pawar, K., Bickham, C. & Lerman, K. Illusions of intimacy: emotional attachment and emerging psychological risks in human-AI relationships. Preprint at https://arxiv.org/abs/2505.11649 (2025). Wang, Y. et al. Evaluating an LLM-powered chatbot for cognitive restructuring: insights from users and mental health professionals. ACM Trans. Comput. Healthcare https://doi.org/10.1145/3820038 (Association for Computing Machinery, 2026). Brandtzaeg, P. B. & Følstad, A. Why people use chatbots. In International Conference on Internet Science (eds Kompatsiaris, I. et al.) 377-392 (Springer, 2017). Lim, M. Y. Memory models for intelligent social companions. In Human-Computer Interaction: The Agency Perspective (eds. Zacarias, M. & de Oliveira, J. V.) 241-262 (Springer, 2012). Pham, C. M., Hoyle, A., Sun, S., Resnik, P. & Iyyer, M. TopicGPT: a prompt-based topic modeling framework. In Proc. 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) (eds Duh, K. et al.) 2956-2984 (Association for Computational Linguistics, 2024). Danaher, J. & McArthur, N. Robot Sex: Social and Ethical Implications (MIT Press, 2017). Purington, A., Taft, J. G., Sannon, S., Bazarova, N. N. & Taylor, S. H. "Alexa is my new bff": social roles, user satisfaction, and personification of the Amazon Echo. In Proc. 2017 CHI Conference Extended Abstracts on Human Factors in Computing Systems (eds Mark, G. et al.) 2853-2859 (Association for Computing Machinery, 2017). Bickmore, T. & Cassell, J. Relational agents: a model and implementation of building user trust. In Proc. SIGCHI Conference on Human Factors in Computing Systems (eds Jacko, J. A. & Sears, A.) 396-403 (Association for Computing Machinery, 2001). Lin, H. & Qiu, L. Sharing emotion on Facebook: network size, density, and individual motivation. In CHI'12 Extended Abstracts on Human Factors in Computing Systems (eds Konstan, J. A. et al.) 2573-2578 (Association for Computing Machinery, 2012). Balani, S. & De Choudhury, M. Detecting and characterizing mental health related self-disclosure in social media. In Proc. 33rd Annual ACM Conference Extended Abstracts on Human Factors in Computing Systems (eds Begole, B. et al.) 1373-1378 (Association for Computing Machinery, 2015). Zhang, R. et al. The dark side of AI companionship: a taxonomy of harmful algorithmic behaviors in human-AI relationships. In Proc. 2025 CHI Conference on Human Factors in Computing Systems (eds Yamashita, N. et al.) 1-17 (Association for Computing Machinery, 2025). Weidinger, L. et al. Ethical and social risks of harm from language models. Preprint at https://arxiv.org/abs/2112.04359 (2021). Ernala, S. K., Burke, M., Leavitt, A. & Ellison, N. B. Mindsets matter: how beliefs about Facebook moderate the association between time spent and well-being. In Proc. 2022 CHI Conference on Human Factors in Computing Systems (eds Barbosa, S. D. J. et al.) 1-13 (Association for Computing Machinery, 2022). Burke, M., Kraut, R. & Marlow, C. Social capital on Facebook: differentiating uses and users. In Proc. SIGCHI Conference on Human Factors in Computing Systems (eds Tan, D. et al.) 571-580 (Association for Computing Machinery, 2011). Yang, D., Yao, Z., Seering, J. & Kraut, R. The channel matters: self-disclosure, reciprocity and social support in online cancer support groups. In Proc. 2019 Chi Conference on Human Factors in Computing Systems (eds Brewster, S. A. et al.) 1-15 (Association for Computing Machinery, 2019).

AI companions may worsen loneliness for vulnerable users

The research highlights risks of relying on AI for companionship and calls for greater public awareness and potential intervention measures. AI companions are becoming increasingly popular, with millions of users forming deep connections with persona-based chatbots and simulated AI partners. But if you're considering leaning on an AI companion for emotional support, a study by researchers in the lab of Diyi Yang, an assistant professor in the Computer Science Department at Stanford, suggests you might want to think twice: As reported in Nature Human Behavior, the researchers found that opening up to chatbots about personal issues made some people feel not better but worse. "While some people turn to chatbots to fulfill social needs, we find that using chatbots in this way doesn't substitute for human connection - in many cases, people actually feel more lonely engaging with AI," says Yang. The study examined the use of AI companions on Character.AI, a generative AI chatbot service where users can design and interact with a wide range of chatbots. Research assistant Yutong Zhang, MS '25, and PhD student Dora Zhao used the research platform Prolific to gather survey data from 1,131 Character.AI users, 244 of whom donated complete chat transcripts. Zhang and Zhao asked participants to identify their primary motivation for using the chatbots (e.g., productivity, entertainment, curiosity, relational) and to describe in their own words the relationships they formed with their AI interlocutors. They then used a collection of AI tools - GPT-4o, LLaMA 3-70B, and TopicGPT - to analyze the data for clues about how users engaged with the chatbots and how those interactions correlated with users' subjective sense of psychological well-being. (The latter was measured using the Comprehensive Inventory of Thriving, a commonly used psychological assessment tool.) The researchers focused on three principal factors: why people were using the chatbots, the intensity of their interactions as measured by things like frequency of use, and how willing people were to share sensitive personal information on topics ranging from emotional distress to substance use and suicidal thinking. They also assessed the size of participants' offline social networks - inquiring, for example, about the number of friends and family members users felt comfortable talking to about personal issues. Interestingly, there was a disparity between how participants described their motives and what their open-ended descriptions and chat transcripts revealed. For example, while just under 12% reported companionship as their primary motive for engaging with the chatbots, slightly over 50% described them using terms like "friend," "companion," or "romantic partner" and more than 80% of the donated chat sessions revolved around seeking emotional and social support from the LLM-powered conversationalists. Social junk food At first blush, more intense usage appeared to be positively associated with well-being. But when the researchers dug into the data, they found that this depended heavily on how and why participants were using the technology. People who interacted intensely and felt pride in their use of the technology or found it meaningful, for example, scored well in terms of subjective well-being. Others, however, did not. The researchers found that intense chatbot use among participants with smaller real-world social networks was correlated with a poor sense of well-being and that association was strongest when companionship was the primary motivation for seeking out the chatbots. Participants who were more willing to share sensitive personal information with their AI companions were also more likely to demonstrate a lower sense of well-being. This was especially striking because self-disclosure tends to have precisely the opposite effect in human relationships. Zhang and Zhao suspect that several factors could be driving these outcomes. Chatbots cannot reciprocally share personal information the way human beings can and may not be able to recognize and respond appropriately to emotionally freighted conversations. At the same time, they are deliberately engineered to keep the interaction going no matter what. "These AI companions are designed to promote engagement," Zhao says. All this makes AI companionship akin to what Zhang calls a "social snack," if not downright junk food: something that offers users an appealing short-term fix for their loneliness and isolation, but which lacks the necessary ingredients for long-term emotional health and well-being. It also risks creating a vicious circle in which people with limited social networks are drawn to intense AI companion use that has the unfortunate effect of further reducing their real-world social contact, leading them to feel even lonelier and more disconnected. Adding guardrails Zhang and Zhao are now trying to figure out which specific features of human-chatbot interaction might be driving the correlation between AI companionship and poor well-being among vulnerable users. Their goal is to identify potential interventions, like imposing usage limits or directing users to real human support when their chat content indicates they need it. In the meantime, they stress the need to educate the public about the potential risks of AI companions, no matter how convenient or attractive they might seem. "We need to make people understand their potential downside, so they'll be more careful about using them," Zhang says.

Study links AI companions to increased loneliness and lower well-being

AI companions are becoming increasingly popular, with millions of users forming deep connections with persona-based chatbots and simulated AI partners. But if you're considering leaning on an AI companion for emotional support, a study by researchers in the lab of Diyi Yang, an assistant professor in the Computer Science Department at Stanford, suggests you might want to think twice: As reported in Nature Human Behavior, the researchers found that opening up to chatbots about personal issues made some people feel not better but worse. "While some people turn to chatbots to fulfill social needs, we find that using chatbots in this way doesn't substitute for human connection - in many cases, people actually feel more lonely engaging with AI," says Yang. The study examined the use of AI companions on Character.AI, a generative AI chatbot service where users can design and interact with a wide range of chatbots. Research assistant Yutong Zhang, MS '25, and PhD student Dora Zhao used the research platform Prolific to gather survey data from 1,131 Character.AI users, 244 of whom donated complete chat transcripts.Zhang and Zhao asked participants to identify their primary motivation for using the chatbots (e.g., productivity, entertainment, curiosity, relational) and to describe in their own words the relationships they formed with their AI interlocutors. They then used a collection of AI tools - GPT-4o, LLaMA 3-70B, and TopicGPT - to analyze the data for clues about how users engaged with the chatbots and how those interactions correlated with users' subjective sense of psychological well-being. (The latter was measured using the Comprehensive Inventory of Thriving, a commonly used psychological assessment tool.) The researchers focused on three principal factors: why people were using the chatbots, the intensity of their interactions as measured by things like frequency of use, and how willing people were to share sensitive personal information on topics ranging from emotional distress to substance use and suicidal thinking. They also assessed the size of participants' offline social networks - inquiring, for example, about the number of friends and family members users felt comfortable talking to about personal issues. Interestingly, there was a disparity between how participants described their motives and what their open-ended descriptions and chat transcripts revealed. For example, while just under 12% reported companionship as their primary motive for engaging with the chatbots, slightly over 50% described them using terms like "friend," "companion," or "romantic partner" and more than 80% of the donated chat sessions revolved around seeking emotional and social support from the LLM-powered conversationalists. Social junk food At first blush, more intense usage appeared to be positively associated with well-being. But when the researchers dug into the data, they found that this depended heavily on how and why participants were using the technology. People who interacted intensely and felt pride in their use of the technology or found it meaningful, for example, scored well in terms of subjective well-being. Others, however, did not. The researchers found that intense chatbot use among participants with smaller real-world social networks was correlated with a poor sense of well-being and that association was strongest when companionship was the primary motivation for seeking out the chatbots. Participants who were more willing to share sensitive personal information with their AI companions were also more likely to demonstrate a lower sense of well-being. This was especially striking because self-disclosure tends to have precisely the opposite effect in human relationships. Zhang and Zhao suspect that several factors could be driving these outcomes. Chatbots cannot reciprocally share personal information the way human beings can and may not be able to recognize and respond appropriately to emotionally freighted conversations. At the same time, they are deliberately engineered to keep the interaction going no matter what. "These AI companions are designed to promote engagement," Zhao says. All this makes AI companionship akin to what Zhang calls a "social snack," if not downright junk food: something that offers users an appealing short-term fix for their loneliness and isolation, but which lacks the necessary ingredients for long-term emotional health and well-being. It also risks creating a vicious circle in which people with limited social networks are drawn to intense AI companion use that has the unfortunate effect of further reducing their real-world social contact, leading them to feel even lonelier and more disconnected. Adding guardrails Zhang and Zhao are now trying to figure out which specific features of human-chatbot interaction might be driving the correlation between AI companionship and poor well-being among vulnerable users. Their goal is to identify potential interventions, like imposing usage limits or directing users to real human support when their chat content indicates they need it. In the meantime, they stress the need to educate the public about the potential risks of AI companions, no matter how convenient or attractive they might seem. "We need to make people understand their potential downside, so they'll be more careful about using them," Zhang says.