The article examines the phenomenon of loneliness in the context of the expanding use of artificial intelligence in students’ everyday communication. The relevance of the study stems from the fact that AI-mediated communication is increasingly perceived not only as a technological service, but also as a form of emotional support capable of influencing the subjective experience of social isolation. The aim of the study was to identify the psychological effects of AI-mediated communication and determine its philosophical implications in the context of students’ experience of loneliness. The study was conducted within the framework of an interuniversity quasi-experiment involving 103 third-year students majoring in Psychology at Astana International University and S. Amanzholov East Kazakhstan University. In the experimental groups, systematic interaction with a generative chatbot was organized over the course of one semester, whereas the control groups studied in the standard mode without a comparable intervention. Data were collected using the UCLA Loneliness Scale and participants’ written reflections. The results showed a statistically significant decrease in the level of subjective loneliness in the experimental groups, whereas changes in the control groups were insignificant. Qualitative analysis identified three stable thematic lines: the perception of AI as an emotionally safe interlocutor, the convenience and accessibility of digital communication, and the recognition of the limitations of AI communication as a substitute for human closeness. It is concluded that AI-mediated communication can perform a compensatory function and temporarily reduce the intensity of loneliness; however, it does not reproduce genuine reciprocity and does not eliminate the need for real interpersonal presence. The practical significance of the study lies in the possibility of using AI as an auxiliary tool for reflection and emotional relief in the university environment.