Synopsis
As artificial intelligence becomes part of everyday university life, students are not just passive users but individuals directly shaped by its decisions. This study explores how psychology students perceive the ethical risks of AI in higher education, focusing on fairness, transparency, data privacy, and academic integrity.
Abstract
Introduction and Positionality
As artificial intelligence (AI) becomes increasingly integrated into higher education, its use in teaching, assessment and administrative processes raises important ethical questions- particularly from the perspective of students, who are directly affected by these systems in their everyday academic lives. Psychology students occupy a particularly interesting position in this context. On the one hand, they are users of AI-driven educational technologies; on the other, their academic training encourages critical reflection on human behaviour, decision-making, power dynamics, and ethical responsibility. This dual perspective has been highlighted in prior research as valuable for interrogating the social and ethical implications of technological systems (Floridi et al., 2018). This study explores psychology students’ perspectives on the ethical risks associated with AI adoption in university settings, with specific attention to issues of fairness, transparency, data privacy, academic integrity, and student agency. By centring student voices, the study aims to contribute to more inclusive and ethically grounded discussions about AI governance in higher education.
Methodology and Theoretical Framework
The study employed an exploratory, survey-based research design informed by ethical AI frameworks (Floridi et al., 2018) and critical perspectives on educational technology (Holmes et al., 2019).. Data were collected through an online questionnaire administered via Google Forms to 200 undergraduate psychology students (N = 200), including 150 from Tbilisi State University and 50 from Alte University.Participation was voluntary and anonymous, and informed consent was obtained prior to survey completion. No personally identifiable information was collected. Beyond measuring general attitudes toward AI, the questionnaire was intentionally designed to capture students’ reflections, concerns and lived experiences with AI-driven academic systems.
The survey consisted of four main sections. First, participants provided basic demographic information and reported their level of familiarity with AI technologies. Second, students were asked about their awareness of common AI applications used in higher education, such as text-matching software, learning analytics systems, automated feedback tools, and online proctoring technologies. Third, the questionnaire addressed ethical dimensions widely discussed in the literature on AI and education, including fairness and bias, transparency and explainability, data privacy, surveillance, and academic autonomy (Floridi et al., 2018; Holmes et al., 2019). Finally, open-ended questions invited participants to express their own views, concerns, and personal opinions regarding the ethical implications of AI use in their academic lives.
Quantitative data were analysed descriptively to identify general trends, while qualitative responses were analysed using reflexive thematic analysis following Braun and Clarke (2006), allowing for an interpretive, experience-oriented exploration of student perspectives to provide greater depth and contextual understanding. Given the exploratory nature of the study, the goal was not statistical generalisation but a nuanced insight into how students interpret and evaluate ethical issues surrounding AI in higher education.
Findings: Fairness, Bias, and Transparency
The findings indicate that psychology students hold complex and often ambivalent attitudes toward AI adoption in universities. Many participants recognised the potential benefits of AI, including increased efficiency, improved accessibility and additional academic support. A majority expressed ethical concerns across multiple dimensions. For example, approximately 70% of participants reported limited understanding of how AI systems function, which was associated with reduced trust. At the same time, these advantages were frequently accompanied by ethical unease. Concerns related to fairness and bias were especially prominent, particularly in relation to AI-supported assessment and automated academic decision-making. Students questioned whether such systems can account for individual circumstances, mental health challenges, or diverse learning trajectories. Several participants also expressed concern that AI-based decisions are often treated as objective or neutral, despite being difficult for students to question, understand, or contest in practice (Floridi et al., 2018).
Transparency emerged as another central theme in students’ responses. 70% of the participants reported limited understanding of how AI systems used by their universities function, what data are collected, and how decisions are generated. This lack of clarity was commonly associated with reduced trust and increased anxiety, especially when AI outputs had direct consequences for grades, evaluations, or academic progression. These concerns closely align with broader critiques of opacity and accountability in AI-driven educational systems (Williamson & Eynon, 2020).
Findings: Privacy, Surveillance, and Human Oversight
Issues related to data privacy and surveillance were also strongly represented. Students frequently described discomfort with continuous monitoring through learning platforms, analytics tools, and online proctoring software, often perceiving these practices as intrusive or disproportionate. Questions surrounding informed consent, data storage and institutional power imbalances were recurrent, echoing ethical challenges previously identified in the learning analytics literature (Slade & Prinsloo, 2013).
Importantly, students tended to distinguish between AI as a supportive tool and AI as a replacement for human judgment. AI applications designed to assist learning, provide feedback, or support academic organisation were generally viewed more positively. In contrast, systems perceived to minimise or remove human oversight - particularly in assessment or disciplinary contexts - elicited stronger resistance and ethical concern.
Conclusion and Implications for Academic Practice
Although some findings align with existing literature, this study contributes by foregrounding student-centred ethical concerns within a specific disciplinary context. The results suggest a gap between institutional AI adoption and student expectations. To address this, universities should: (1) implement transparent AI disclosure policies; (2) ensure human-in-the-loop decision-making in high-stakes contexts; and (3) actively involve students in the co-design of ethical guidelines and governance frameworks.
By centring student voices, this study highlights the need for more participatory, transparent, and ethically grounded approaches to AI integration in higher education. Actively involving students not only as users but as stakeholders in ethical discussions may help universities foster more responsible, trustworthy and socially legitimate AI practices in higher education (Holmes et al., 2019; Williamson & Eynon, 2020).
Keywords: artificial intelligence; ethics in higher education; student perspectives; academic integrity; data privacy; psychology students; reflexive thematic analysis.
References:
Braun, V., & Clarke, V. (2006).
Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa
Floridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., … Vayena, E. (2018).
AI4People-An ethical framework for a good AI society: Opportunities, risks, principles, and recommendations. Minds and Machines, 28(4), 689–707. https://doi.org/10.1007/s11023-018-9482-5
Holmes, W., Bialik, M., & Fadel, C. (2019).
Artificial intelligence in education: Promises and implications for teaching and learning. Center for Curriculum Redesign.
Slade, S., & Prinsloo, P. (2013).
Learning analytics: Ethical issues and dilemmas. American Behavioral Scientist, 57(10), 1510–1529. https://doi.org/10.1177/0002764213479366
Williamson, B., & Eynon, R. (2020).
Historical threads, missing links, and future directions in AI in education. Learning, Media and Technology, 45(3), 223–235. https://doi.org/10.1080/17439884.2020.1798995