Recent research highlights the rapidly growing integration of artificial intelligence (AI) in higher education, emphasizing its potential to support learning, assessment, and instructional design (Bond et al., 2024). At the same time, empirical studies increasingly point to persistent challenges related to pedagogical readiness, ethical use, data protection, unequal access, and broader social and developmental implications of AI-supported learning environments (Holmes et al., 2021; Nguyen et al., 2022). These challenges are particularly salient in teacher education, where future teachers’ competencies, beliefs, and ethical orientations play a crucial role in shaping classroom practices and long-term educational outcomes (Su & Yang, 2023).

The conceptual framework of this study is grounded in interdisciplinary research on technology integration in education, teacher professional competence, and responsible AI use. It draws on the Technological Pedagogical Content Knowledge (TPACK) framework (Voogt et al., 2012), which conceptualizes effective technology integration as the interplay between technological, pedagogical, and content knowledge. In addition, the framework incorporates emerging perspectives on ethical and responsible AI in education, including community-wide ethical frameworks (Holmes et al., 2021), pedagogical and policy-oriented analyses of AI risks (Nguyen et al., 2022), and the European Digital Competence Framework for Citizens (DigComp 3.0), which explicitly addresses ethical, critical, and responsible use of digital technologies (European Commission, Joint Research Centre [JRC], 2025).

The Slovenian context of teacher education is relatively unified and small compared to many other EU countries. Initial teacher education is offered by only three universities and five faculties nationwide, with study programmes predominantly at the Master’s level and some at the Bachelor’s level. Structurally, Slovenian teacher education aligns with the Bologna two-cycle model and is therefore comparable to most EU systems in terms of qualification level and formal organization (European Commission/EACEA/Eurydice, 2023). Despite this structural alignment, comparative European indicators suggest challenges relevant to AI integration. While Slovenia maintains relatively high overall educational attainment, broader indicators such as population-level digital skills remain below the EU average, which may indirectly affect preparedness for advanced digital and AI-related pedagogies (European Commission, Directorate-General for Education, Youth, Sport and Culture, 2025). Moreover, OECD evidence indicates that recent graduates in Slovenia rate the quality of their initial teacher education below the OECD average, suggesting room for improvement in areas such as pedagogical innovation and digital competence development (OECD, 2016; OECD, 2025).

While existing international literature has examined attitudes toward AI adoption in education, fewer studies have systematically compared teachers’ and students’ perceptions within teacher education programmes, particularly across both general pedagogical challenges and ethical challenges of AI use within a shared institutional context. Quantitative evidence comparing these stakeholder groups remains limited, especially in smaller educational systems that nevertheless reflect broader EU policy priorities and structural models.

The objective of this study is therefore to examine how teachers and students in Slovenian teacher education programmes differ in their perceptions of AI-related challenges, and to situate these findings within a wider European discourse on responsible AI integration in higher education. The study addresses the following research question: What differences exist between teachers’ and students’ perceptions of (a) general pedagogical challenges and (b) ethical challenges related to the use of artificial intelligence in teacher education programmes in higher education?

By comparing teachers’ and students’ perceptions across these dimensions, the study contributes empirical evidence to European and international debates on AI in teacher education. The findings aim to inform professional development, institutional policy, and the responsible, pedagogically grounded integration of AI in higher education in pedagogical study programmes on national and international level.

Research design The study employed a quantitative research design using a cross-sectional survey to examine teachers’ and students’ perceptions related to the use of artificial intelligence (AI) in teacher education. This study was conducted as part of the research project “Generative Artificial Intelligence in Education” (Project code: NRP 3350-24-3502). The project is funded by the Republic of Slovenia, the Ministry of Education, and the European Union – NextGenerationEU. Participants The student sample (n = 460) consisted predominantly of students enrolled in first-cycle (Bachelor’s) university programmes (67%), followed by Master’s-level students (16.3%) and students in professional higher education programmes (14.3%) and PhD programmes (2.3%). Geographically, students were recruited from three Slovenian institutions educating future teachers: two faculties at the University of Ljubljana (50.0%), the Faculty of Education, University of Maribor (25.0%), and the Faculty of Education, University of Primorska (25.0%). The teacher sample (n = 119) included higher education teachers and teaching assistants. Most participants were female (71.3%), followed by male participants (27.8%); one participant (0.9%) did not report gender. The average age of teachers was 46.35 years, with an average of 20.59 years of professional experience. Their institutional distribution was similar that of the student sample. Instrument Data were collected using a questionnaire, designed specifically for this study. The instrument was developed through multiple expert reviews conducted by members of the research team and external experts. This paper reports only results related to general pedagogical challenges and ethical challenges of AI use. General challenges included self-assessed AI knowledge, pedagogical readiness of teachers and students, effective ability to use AI, digital inequality and access, overreliance on AI, data privacy protection, and potential social and emotional impacts (e.g. interaction patterns, motivation, and emotional engagement). Ethical challenges included plagiarism, intellectual property, bias and unfairness, data privacy concerns, false or fabricated information, anthropomorphization of AI tools, environmental impacts, social impacts, dependence on AI use, and lack of critical judgment. Internal consistency for the teacher questionnaire was acceptable for exploratory research (general pedagogical challenges: Cronbach’s α = .68; ethical challenges: α = .86). For students, reliability was slightly higher (general pedagogical challenges: α = .69; ethical challenges: α = .89). Given the newly developed nature of the instrument, these values were considered satisfactory. Data collection and ethics Data were collected between December 2024 and January 2025. The study adhered to fundamental research ethics principles. Ethical approval was obtained from the Research Ethics Committee of the Faculty of Arts, University of Maribor (No. 038-30-196/2024/42/FF/UM).

Independent-samples t-tests revealed statistically significant differences between teachers and students across all examined variables of general pedagogical challenges related to AI use in teacher education. Teachers reported higher self-assessed knowledge of AI than students and expressed stronger agreement with statements concerning pedagogical readiness, unequal access to AI tools, unintended negative effects of AI use, data privacy challenges, and potential social and emotional impacts on learners. Effect size for general pedagogical challenges estimates (Cohen’s d = 0.26–0.56) indicate small to moderate practical significance, with the largest differences observed for pedagogical readiness, access inequalities, and data privacy concerns. These findings suggest that teachers approach AI integration with a more cautious and risk-aware perspective, likely reflecting their professional responsibility for pedagogical quality, ethical compliance, and student well-being. Teachers also reported significantly higher concern than students across most ethical challenges of AI use, including plagiarism, intellectual property, bias and unfairness, data privacy, false or fabricated information, anthropomorphization of AI tools, environmental impacts, social impacts, and lack of critical judgment (p ≤ .002). No statistically significant difference was found for dependence on generative AI use, indicating a shared perception between the two groups on this issue. Effect sizes for ethical challenges ranged from small to large (Cohen’s d = 0.15–0.79). The strongest effects were observed for plagiarism and intellectual property, highlighting substantially greater concern among teachers. Moderate effects emerged for bias and unfairness, data privacy, misinformation, and lack of critical judgment, while smaller effects were found for anthropomorphization and environmental and social impacts. The findings indicate systematic differences between teachers’ and students’ perceptions of both general and ethical challenges of AI use in teacher education. Teachers consistently assign greater importance to AI-related risks, underscoring the need for targeted professional development, ethically grounded AI integration, and curricular approaches that address both technical competence and critical, responsible AI use in higher education.

Bond, M., Khosravi, H., De Laat, M., Bergdahl, N., Negrea, V., Oxley, E., Pham, P., Chong, S. W., & Siemens, G. (2024). A meta systematic review of artificial intelligence in higher education: A call for increased ethics, collaboration, and rigour. International Journal of Educational Technology in Higher Education, 21, Article 4. https://doi.org/10.1186/s41239-023-00436-z European Commission, Directorate-General for Education, Youth, Sport and Culture. (2025). Education and training monitor 2025. Publications Office of the European Union. https://doi.org/10.2766/2720910 European Commission, Joint Research Centre. (2025). DigComp 3.0: The Digital Competence Framework for Citizens. Publications Office of the European Union. https://doi.org/10.2760/144121 European Commission/EACEA/Eurydice. (2023). Initial education for teachers working in early childhood and school education: Slovenia. Eurydice national education systems. https://eurydice.eacea.ec.europa.eu/eurypedia/slovenia/initial-education-teachers-working-early-childhood-and-school-education Holmes, W., Porayska-Pomsta, K., Holstein, K., Sutherland, E., Baker, T., Buckingham Shum, S., Santos, O. C., Rodrigo, M. T., Cukurova, M., Bittencourt, I. I., & Koedinger, K. R. (2021). Ethics of AI in education: Towards a community-wide framework. International Journal of Artificial Intelligence in Education, 32(2), 504–526. https://doi.org/10.1007/s40593-021-00239-1 Nguyen, A., Ngo, H., Hong, Y., Dang, B., & Nguyen, B. (2022). Ethical principles for artificial intelligence in education. Education and Information Technologies, 28, 4221 - 4241. https://doi.org/10.1007/s10639-022-11316-w. OECD. (2016). Education policy outlook: Slovenia. OECD Publishing. https://www.oecd.org/content/dam/oecd/en/about/projects/edu/education-policy-outlook/398027-Education-Policy-Outlook-Country-Profile-Slovenia.pdf OECD (2025). Education at a glance 2025: OECD indicators. OECD Publishing. https://doi.org/10.1787/1c0d9c79-en Su, J. H., & Yang, W. P. (2023). Unlocking the power of ChatGPT: A framework for applying generative AI in education. ECNU Review of Education, 6(3), 355–366. https://doi.org/10.1177/20965311231168423 Voogt, J., Fisser, P., Pareja Roblin, N., Tondeur, J., & van Braak, J. (2012). Technological pedagogical content knowledge – A review of the literature. Journal of Computer Assisted Learning, 29(2), 109–121. https://doi.org/10.1111/j.1365-2729.2012.00487.x