JMIR Publications

ABSTRACT

Background:

The rapid integration of artificial intelligence(AI) into everyday life haw introduced new ways of interaction between human and technology, particularly through AI-powered chatbots. These tools are used by Generation Z, a cohort already identified as vulnerable to loneliness and social isolation. However, limited research has examined the psychosocial implications of AI chatbot use in this population.

Objective:

This study aimed to investigate the relationship between AI chatbot use and psychosocial outcomes, specifically loneliness and social isolation among Generation Z.

Methods:

A cross-sectional quantitative survey was conducted with 472 participants aged 15-30 years. Participants completed validated instruments: the General Attitudes toward AI Scale(GAAIS), the UCLA Loneliness Scale and Lubben Social Network Scale(LSNS). Demographic information and frequency/motives of AI chatbot use were also collected. Descriptive statistics, correlation analyses, and regression models were used to assess associations between chatbot use, loneliness, and social isolation.

Results:

The final sample consisted of 472 participants (N=472). The demographic composition of the sample included 283(59.96%) females and 189 males (40.04%), with a mean age of 22,54(SD=3,528) (Table 1). Most participants reported regular use of smartphones and digital communication tools. Regarding educational background, 70.8% had current or completed tertiary education, 21.4% were high school graduates, and 7.8% had vocational or technical training. In terms of chatbot usage, 51.5% of participants were regular users (daily or frequently), while only 7.2% reported no use. The mean loneliness score (UCLA Scale) was 42.84 (SD = 11.36), indicating moderate loneliness, while the mean LSNS score for social isolation was 37.99 (SD = 8.97), suggesting relatively intact social networks. The results also indicate that while the frequency of chatbot use is not significantly associated with either perceived loneliness or social isolation, the underlying motivations for use play a critical role. Specifically, emotional motives-such as the need to talk or to address personal concerns-were positively associated with increased levels of loneliness, yet not with greater social isolation. This suggests that individuals who turn to chatbots for emotional reasons may experience a subjective sense of loneliness despite maintaining objectively adequate social networks.

Conclusions:

This study examined the interplay between AI chatbot use and psychosocial outcomes-specifically loneliness and social isolation-among Generation Z. By moving beyond usage frequency and investigating the motivations underpinning chatbot interaction, the study offers important insights into how digital tools are integrated into the emotional and social lives of young people. The findings demonstrate that usage frequency alone does not predict levels of loneliness or social isolation. Rather, the underlying purpose of use emerges as a significant factor. These results highlight the need to account for contextual and motivational variables when evaluating the mental health implications of AI tools. As conversational agents become more embedded in daily routines and social practices, researchers, designers, and policymakers should consider not only the frequency of engagement but also the underlying psychological drivers. While AI chatbots may provide short-term emotional relief and a sense of companionship, they cannot substitute for the depth and reciprocity of meaningful human relationships. Their role should be framed not as replacements for social connection, but as complementary tools that must be integrated thoughtfully and ethically into a broader system of mental health and social support. Clinical Trial: Not applicable.