Date
Publisher
arXiv
Artificial intelligence (AI)-based learning assistants and chatbots are increasingly integrated into higher education. While these tools are often evaluated in terms of technical performance, their successful and ethical use also depends on psychological factors such as trust, perceived risk, technology anxiety, and students' general attitudes toward AI. This paper adopts a psychology-oriented perspective to examine how university students form trust in AI-based learning assistants. Drawing on recent literature in mental health, human-AI interaction, and trust in automation, we propose a conceptual framework that organizes psychological predictors of trust into four groups: cognitive appraisals, affective reactions, social-relational factors, and contextual moderators. A narrative review approach synthesizes empirical findings and derives research questions and hypotheses for future studies. The paper highlights that trust in AI is a psychological process shaped by individual differences and learning environments, with practical implications for instructors, administrators, and designers of educational AI systems.
Study at a glance
- Question: Which psychological factors shape university students' trust in AI-based learning assistants?
- Setting and data: Peer-reviewed studies on university students and young adults published between 2018 and 2025.
- Method: Narrative literature review and conceptual framework development.
Why it matters
While educational AI tools are often evaluated on technical performance, successful use depends on psychological factors. Frameworks are needed to understand how students trust AI and avoid underuse or harmful overreliance.
Key findings
- Higher cognitive appraisals, such as perceived competence, accuracy, and usefulness, are consistently associated with increased trust, continuance intention, and perceived learning support.
- Technology anxiety is associated with lower trust and reduced usage, whereas feelings of emotional safety relate to greater student engagement and willingness to experiment with AI.
- Supportive social cues, perceived empathy, and autonomy support relate to appropriate reliance, while transparent communication and clear policies weaken the negative impact of privacy concerns.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Why use AI?
Study design
