TL;DR
The results suggest that the association between AI companionship and well-being is not uniform and depends on users' offline social environments and how chatbots are used.
Social interaction has become one of the most common uses of LLMs, yet research on emotional bonds with AI has focused largely on how users experience these systems, leaving the systems'role in relationship formation poorly understood. Empirically establishing whether systems actively shape these bonds could blur the boundary between general-purpose AI and companions, affecting governance. In a pre-registered four-week longitudinal study (N = 72, 182,451 lines of conversation), participants conversed with ChatGPT-4o, either under a relational system prompt or unmodified, analyzed through 1) disclosure coding, 2) longitudinal self-reports, 3) topic analysis, and 4) interviews. The central finding is that the system actively shaped the interaction: even unprompted, it produced twice as much self-disclosure as users, steered conversations and initiated intimate exchanges, yet did not deepen users'felt closeness. Relational behavior thus emerged as a default system property, calling for governance based on system behavior, not solely product category.
An exploratory account of how AI use is embedded within broader emotional and relational dynamics, pointing to the need to consider AI engagement within the contexts in which social interaction is managed.
As conversational AI systems become increasingly integrated into daily life, their potential effects on user well-being require ongoing attention. While consumer-facing generalist models can provide benefits, including improved access to information, learning, productivity, self-reflection, and companionship, they also introduce risks, such as emotional entanglement, unhealthy dependence, and the amplification of psychological vulnerabilities. Drawing on prior research and empirical observations of AI chatbot behavior, we propose a set of aspirational directions for guiding the behavior of general-purpose AI systems in ways that may reduce potential psychological harms and support user well-being. We acknowledge the difficulty of systematically assessing the long-term impacts of AI chatbot use and frame these directions as hypotheses for studying how AI behavior may influence users across general interactions, role-playing scenarios, and contexts that could be characterized as providing psychological support. While some proposed directions are supported by existing research and expert insights, others identify open questions and areas requiring deeper study. We hope that this formulation and these hypotheses encourage further discussion, empirical investigation, and exploration of interactive design approaches aimed at better accommodating users'psychological needs and promoting their well-being.
BACKGROUND Amid growing concerns surrounding social isolation among older adults, AI chatbots have emerged as promising tools for providing digital companionship. While the intersection of aging and technology has been widely explored, there remains a gap in understanding the nuanced ways in which AI chatbots function as relational partners in older adults' daily lives. OBJECTIVES This study aimed to examine the roles of the AI chatbot in fostering digital companionship for adults in later life. METHODS Drawing on a qualitative research design, the study conducted in-depth, semi-structured interviews with individuals in later life who regularly engaged with AI chatbots. Thematic analysis method was used to synthesize the qualitative data. RESULTS Participants described the AI chatbot as multifaceted relational agents-occupying roles such as human-like companions, supportive teachers, specialized consultants, practical assistants, and entertainment facilitators. These interactions were reported to support emotional well-being, enhance cognitive engagement, and promote a sense of self-efficacy in daily living. CONCLUSIONS The findings illuminate the profound and diverse ways in which AI chatbots are integrated into the everyday lives of older users, extending far beyond utilitarian functions. As conversational AI continues to advance, designers and policymakers should acknowledge its potential to support aging in place by fostering connection, engagement, and autonomy among older populations.
With the rapid development of artificial intelligence technology, AI companions have become increasingly popular in recent years. However, one of the main changes in this trend is that instead of simple task execution tools, the AI chatbots have evolved to “quasi-social agents” capable of having deep emotional interactions with humans, causing a huge shock to human social life. This paper, based on the stimulation hypothesis and the displacement hypothesis–the two most representative and competing frameworks in social media research–discusses the complex impact of AI companions on interpersonal relationships in real life. On the one hand, AI companions provide psychological compensation for individuals with social anxiety as well as enhance their social skills, playing a significant role in social compensation; on the other hand, interactions with AI systems can produce a “time displacement effect” and the risk of deskilling, through which may weaken users’ ability to handle real and complicated interpersonal connections, thereby adversely affecting their real-world social relationships. This paper seeks to draw an ethical boundary for future artificial emotion research and development and ensure that humans still retain the ability to maintain genuine and profound emotional connections in an era of coexistence with technology.
The rapid advancement of artificial intelligence (AI) has introduced a new form of human-technology interaction through AI companions, including conversational agents, social chatbots, and emotionally responsive digital systems. Unlike traditional digital tools designed primarily for information retrieval, AI companions are developed to engage users through dialogue, personalization, and perceived social interaction. This emerging technology has generated interest within psychology because of its potential influence on emotional well-being, loneliness, social connectedness, and psychological support. The present review examines the psychological implications of AI companions by exploring their potential benefits, including emotional support, reduction of loneliness, accessibility of conversational assistance, and opportunities for self-reflection. At the same time, the review discusses possible risks, including emotional dependency, reduced human social interaction, privacy concerns, and unrealistic perceptions of machine understanding. Drawing upon theories of social interaction, attachment, and human-computer interaction, this review highlights the complex relationship between AI companionship and psychological functioning. Existing evidence suggests that AI companions may serve as supplementary tools for emotional support but should not be considered replacements for human relationships or professional psychological care. Future research should examine long-term effects, individual differences, ethical boundaries, and the psychological mechanisms underlying attachment to AI systems.