TRUST IN STUDENTS’ VOLUNTARY USE OF CHATGPT AND GENERATIVE AI IN HIGHER EDUCATION: A SYSTEMATIC LITERATURE REVIEW

Abstract

The rapid diffusion of generative artificial intelligence (GenAI), particularly tools such as ChatGPT, has intensified students’ voluntary reliance on systems that operate under epistemic uncertainty. While existing research on GenAI adoption predominantly emphasises behavioural intention and usage outcomes, the psychological role of trust in enabling reliance remains conceptually fragmented. This systematic literature review synthesises empirical evidence on trust and techno-trust in students’ voluntary use of GenAI within higher education contexts. Guided by PRISMA 2020, a systematic search of Scopus and Web of Science identified 11 empirical studies published between 2023 and 2025 for descriptive synthesis. The findings reveal substantial heterogeneity in how trust is defined, operationalised, and positioned within empirical models. Trust is predominantly conceptualised as a cognitively oriented evaluation of system reliability or output credibility. Across studies, it is variably positioned as an antecedent, mediator, moderator, outcome, or remains implicitly embedded within adoption constructs. Behavioural intention and use-related outcomes dominate the literature, whereas reliance, acceptance, and calibrated trust receive comparatively limited empirical attention. By consolidating fragmented evidence through an integrative typology and analytical mapping of trust positioning, this review clarifies the inconsistent analytical roles assigned to trust in voluntary GenAI use. The findings reconceptualise trust as a psychological reliance mechanism operating under epistemic risk rather than a peripheral adoption variable. Educationally responsible engagement with GenAI therefore depends not on maximising trust, but on cultivating calibrated and warranted reliance. This synthesis provides a clearer foundation for future research on trust calibration, epistemic judgement, and decision-making in the educational use of generative artificial intelligence.

Ähnliche Arbeiten