Generative Artificial Intelligence (GenAI) is rapidly transforming societies and economies. It promises considerable benefits for global challenges, but  also presents substantial risks especially about mis- and dis-information (OECD, 2024). In an educational context, the discussion proceeds along similar lines, usually based on the potential impact of GenAI on teaching and learning (see, for instance, Unesco or OECD reports such as the one from Miao & Holmes, 2023 or OECD, n.d.). The perceived impact can, on the one hand, lead to unbridledly being convinced of benefits and therefore ‘completely embracing’ the use of GenAI. On the other hand, the perceived impact can also result in being convinced of doomsday scenarios and ‘completely banning’ the use of GenAI in education (Walter, 2024). The discourse about GenAI in education generally takes the form of speculations and visions about the future. For instance, Microsoft (in Helmond, 2024) asks to “Imagine a future where every student has a personalized learning path, where faculty can focus on teaching instead of administrative tasks, and where academic research accelerates breakthrough discoveries. This is not a distant vision – generative AI is making it possible today.” According to Wiliamson (2024), these visions are accompanied by terms like personalized learning, customization, 24/7 availability, and improved learning. Those opposed to GenAI stress the potential risks in terms like bias, incorrect information, de-contextualization, technology dependency, integrity, and loss of autonomy. In practice, the discussion seems to produce a spectrum of stances: from ignorance towards rethinking education and everything in between (Lodge et al., 2023).

Of course, speculations like this and descriptions of GenAI’s potential for education are more and more underpinned by research. From their meta systematic review, Bond et al. (2024) found that research is mostly about GenAI in general and is focused on the teaching tasks. And surely, there are many teachers who experiment with specific GenAI applications for parts of their teaching. Also, various studies indicate that GenAI provide opportunities for students for, for example, personalized learning (Crompton & Burke, 2024). However, do students really take advantage of these opportunities? Does it help their learning, or does it lead to outsourcing difficult tasks? Anyway, it is crucial that students use GenAI in a responsible and critical manner and that is a point to note (Bond et al., 2024).

We know that the degree to which students accept new technologies, predicts whether they will actually use the technology. As Shoufan (2023) states, students easily accept and get used to new technologies. Together with good availability of the technology, this probably leads quickly to fixed habits of GenAI use. So, we know what could be possible uses (Bond et al., 2024), we also know how students perceive GenAI in a special situation set up for this purpose (e.g., Shoufan, 2023), but we barely know what students actually use when they themselves are busy with their studies. More insights are needed when we aim at preparing students for deliberately using GenAI. Indeed, several studies (e.g., Bond et al., 2024; Shoufan, 2023; Walter, 2024) call for research that dives deeper into students’ perceptions of their use of GenAI. This study addresses this gap by focusing on the student: what do they use, for what, and why (not)?

A mixed-methods design was set up in order to conduct an exploratory study on the question What GenAI tools do students in Higher Education use for what, and why (not)? First, a quantitative questionnaire was developed, containing questions about tools, tasks, and reasons for using these tools (or not). The parts about tasks and reasons exist of several predefined statements and also an open question in which respondents can make their own additions. Table 1 contains example questions for each part of the questionnaire. The questionnaire was spread in November 2024 among the students from an educational and an economic program of a University of Applied Sciences in the Netherlands. Students answered anonymously (n=93). Table 1. Example questions Focus on Example (answer options) Tools Can you name some tools you have used more than once? (open) Tasks Please tick the box that indicates for which of the following categories you use GenAI for: - creative and visual applications, e.g. generation of images, visuals or video etc. (5-point Likert scale: no use – several times/year – several times/month – once/ week – several times/ week) Reason for use Why are you using GenAI applications? - time savings, e.g. the use of GenAI helps me complete assignments faster. (5-point Likert scale: totally disagree > totally agree) Reason for non-use Are the examples below a reason for you not to use GenAI? - data security, for example, because I don't know what GenAI does with non-use my data. (5-point Likert scale: totally disagree > totally agree) Using an interview guideline, the quantitative date are enriched by semi-structured questions with which we got a more specific portrait of how students use GenAI. Students (n=11) volunteered for this one-hour interview, which takes place in February 2025. The interviews will be recorded and summary transcribed. The summary shall be given to the participants for verification. Descriptive statistics were carried out and correlations were calculated for the quantitative data. Reliability of the questionnaire proved good (Cronbach’s α = .82). We will analyze the qualitative data intuitively, searching for deepening the patterns that the quantitative analysis yielded. At the time of this proposal, the quantitative data have been analyzed descriptively. During the conference, we will also present calculated correlations as well as the results of the qualitative analysis.

A majority (>71%) of the students use GenAI, although younger students (<25 years) significantly more often (92%) than older students (about 70%). ChatGPT is by far the most used tool, followed by Co-pilot. GenAI is used to generate and process texts and to gather and process information (several times per week/month). Students use GenAI less for reinforcing learning (e.g. understanding concepts better, preparing for a test) (several times per month). For analytical and computational tasks, programming and computer simulations and creative and visual applications, GenAI is not or barely used. Users of GenAI do so because of ease of use (78%) and time savings (62%). Preparation for future work and contribution to the quality of one's work seem hardly reason for use. 14% of the students do not use GenAI, of which three quarters are in the early stages of study. Students in the beginning phase of their study give reasons that seem to be related to uncertainty, namely not knowing what works, not knowing what is allowed, or afraid of not learning anything. In the graduation phase, students more often mention reliability of output as a reason, which they can assess better because they logically have a broader knowledge base than students in previous phases. Overall, ethical considerations (such as academic integrity, bias, and reliability) hardly seem to play a role. From the interviews, we expect that the reasons for (not) using GenAI can be substantiated with contextualization, more specific arguments and nuances. We also expect to be able to say more precisely how students work with GenAI, including what prompts they give and what they do with the output. As a preliminary conclusion, it seems the majority of students use GenAI only for a limited number of tasks and because it is easy to use and saves time.

Bond, M., Khosravi, H., De Laat, M., Bergdahl, N., Negrea, V., Oxley, E., ... 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, (2024) 21:4. https://doi.org/10.1186/s41239-023-00436-z Crompton, H., & Burke, D. (2024). The educational affordances and challenges of ChatGPT: State of the field. Techtrends, 68:380-392. https://doi.org/10.1007/s11528-024-0093-0 Johnston, H., Wells, R.F., Shanks, E.M., Boey, T., & Parsons, B. N. (2024). Students’ perspectives on the use of generative artificial intelligence technologies in higher education. International Journal for Educational Integrity, (2024)2012, pp. 2-21. Lodge, J.M., Howard, S., & Broadbent, J. (2023, May 1). Assessment redesign for generative AI: A taxonomy of options and their viability. [Post]. https://www.linkedin.com/pulse/assessment-redesign-generative-ai-taxonomy-options-viability-lodge/ Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. Unesco. OECD. (2024). OECD Digital Economy Outlook 2024 (Volume 1): Embracing the Technology Frontier. OECD Publishing. https://doi.org/10.1787/a1689dc5-en OECD. (n.d.). Artificial intelligence and education and skills. Retrieved November 14th, 2024, from https://www.oecd.org/en/topics/sub-issues/artificial-intelligence-and-education-and-skills.html Shoufan, A. (2023). Exploring students’ perspectives of ChatGPT: Thematic analysis and follow-up study. IEEE Education Society Section, 11, pp. 38805-38818. https://doi.org/0.1109/ACCESS.2023.3268224 Vo, A., & Nguyen, H.(2024). Generative Artificial Intelligence and ChatGPT in Language Learning: EFL Students' Perceptions of Technology Acceptance. Journal of University Teaching and Learning Practice, 21(6). Walter, Y. (2024). Embracing the future of Artificial Intelligence in the classroom: the relevance of AI literacy, prompt engineering, and critical thinking in modern education. International Journal of Educational Technology in Higher Education, 21,15 (2024). https://doi.org/10.1186/s41239-024-00448-3 Williamson, B. (2024). AI in schools. Keywords of a public problem. [Presentation]. GDS Kennisnet symposium, Netherlands, Utrecht, 2024, November 7.