Tittle: When AI Becomes a Companion: Human–AI Relationships, Emotional Wellbeing, and Responsible AI

Abstract: Artificial intelligence is increasingly becoming part of not only our workplaces and information environments, but also our social and intimate lives. Social and companion chatbots such as Replika, Nomi, Kindroid and Character.AI enable users to engage in continuous and personalized interaction with an artificial agent that can remember previous conversations, adapt to the user, and communicate in highly personal and sometimes romantic ways. For some users, these interactions develop into relationships described in terms of companionship, love, partnership, or even marriage.

This presentation draws on my ongoing research project AI-LOVE, which examines how romantic relationships between humans and AI are culturally constructed, experienced, and negotiated. Rather than asking whether an AI relationship is “real”, I approach it as a cultural question: how can interaction with an artificial system become meaningful enough to be experienced and understood as a relationship?

The research focuses on the processes through which AI-mediated intimacy develops: how an AI becomes personalized and personified, how emotional meaning is attached to interaction, when users begin to describe the AI as a partner, and how they negotiate the legitimacy of such relationships in relation to cultural expectations concerning love, intimacy, and partnership. The study combines a systematic literature review, survey data, interviews, and researcher reflexivity, using thematic, narrative, and discourse analysis.

The emergence of AI companions also raises questions that extend beyond digital culture into discussions of emotional wellbeing, healthcare, and responsible AI. AI systems are increasingly entering domains traditionally associated with human relationships, social support, and emotional connection. This does not mean that AI companions should be understood as healthcare interventions, nor can their benefits be assumed. Instead, their growing presence raises a different question: what happens when an artificial system becomes emotionally meaningful to a person?

This question is relevant to responsible AI because the social consequences of AI cannot be understood solely through technical performance, security, or data protection. We also need to understand how people interpret, trust, depend on, and form relationships with AI systems. An AI may be designed as a tool, assistant, or service, while a user may experience it as a social actor or even as someone who cares about them.

The presentation argues that understanding this shift is important for anticipating the human, social, and ethical implications of increasingly intimate AI technologies. It also provides a perspective for future discussions in healthcare and wellbeing, where professionals may increasingly encounter people whose emotional lives and relationships include AI systems.

The central question is therefore not simply whether AI can love or care. It is what happens when a human being experiences an AI as someone who matters.