ARTIFICIAL INTELLIGENCE AS A DIGITAL INTERLOCUTOR: TRANSFORMING INTERPERSONAL INTERACTION IN CONTEMPORARY KAZAKHSTAN

Abstract

The rapid integration of generative artificial intelligence into everyday life is transforming not only information-seeking and professional practices but also interpersonal communication, advice-seeking, self-disclosure, and emotional support. This study examines how generative AI functions as an adviser, conversational partner, and communicative resource among adults in Almaty, Kazakhstan, focusing on two potentially coexisting processes: interaction substitution, whereby AI partially replaces human interaction in selected communicative and supportive functions, and interaction augmentation, whereby AI facilitates subsequent communication with other people. A quantitative cross-sectional survey was conducted among 427 adults aged 18 years and older. The analysis included descriptive statistics, reliability analysis, Spearman correlations, group comparisons, and multiple regression modelling. The findings showed that interaction augmentation was more prevalent than interaction substitution. Personal AI use was strongly associated with both substitution and augmentation, while substitution was additionally associated with greater AI self-disclosure and loneliness and with lower perceived social support. AI trust emerged as the strongest independent correlate of interaction augmentation. Significant age differences were also identified, with younger respondents and students reporting substantially higher levels of both forms of AI-mediated interaction. Importantly, substitution and augmentation were positively correlated, indicating that they represent coexisting rather than mutually exclusive patterns of human–AI interaction. The findings suggest that generative AI does not simply replace interpersonal communication but contributes to a broader redistribution of communicative functions between human and technological actors. By providing evidence from Almaty, the study extends the emerging literature on human–AI interaction to a Central Asian urban context and highlights the need to conceptualise generative AI as an increasingly embedded component of contemporary social interaction