Features of social perception of an algorithmised psychotherapist by Ukrainian large language models (LLMs) users

The growing demand for mental health support, along with the development of artificial intelligence technologies, has led to the increasing use of large language models (LLMs) for obtaining psychological assistance. The aim of this study was to examine the social perception of generative artificial intelligence in the role of a psychotherapist among Ukrainian users of large language models. The survey questionnaire included a sociodemographic section as well as questions regarding respondents’ experience of receiving psychological assistance through generative artificial intelligence (AI). In addition, a projective colour association technique and a projective method of selecting an associative image were employed in order to identify how users of an algorithmised psychotherapist perceive it. The final sample consisted of 85 participants aged between 17 and 54, including 22 men and 58 women. The study showed that generative artificial intelligence in the role of a psychotherapist is perceived as calm, unemotional, genderless, and cold. It was also found that an algorithmised psychotherapist based on LLMs is simultaneously perceived as a coherent structure and as something mysterious and hidden. At the same time, generative AI in the role of a psychologist is not associated with evil or deceit, which may indicate a potential vulnerability to confident error and therefore requires further investigation. It was determined that perceptions of generative artificial intelligence do not depend on age or prior experience of receiving professional psychological assistance. The visual representation of generative AI is influenced by mass media in terms of colour, but not in terms of anthropomorphisation, provided that usage remains moderate. The results may be used by psychologists, AI developers, and policymakers to design ethically grounded and socially sensitive AI-based mental health tools