Conference Agenda

The present contribution examines the usage of generative AI by pre-service teachers for learning purposes during their teacher education studies. The study is grounded in a constructivist understanding of learning, according to which learning is understood as an active, constructive, and self-regulated process. From this perspective, learning does not occur through the purely passive reception of knowledge but requires active engagement with new information (Schunk, 2012). In this process, the learner, together with their prior knowledge, experiences, and social interactions, is placed at the center of learning (Schunk, 2012; Vygotsky, 1978). Closely linked to this understanding of learning is the theoretical concept of self-regulated learning, which emphasizes the learning objective as well as the monitoring and control of cognition, motivation, and behavior during the learning process (Pintrich, 2000). Based on Zimmermann’s (2000, p. 16f.) distinction between „forethought“, „performance or volitional control“, and „self-reflection“, Schmitz and Schmidt (2007) conceptualize self-regulated learning as a process consisting of a pre-action, action, and post-action phase. Recent approaches have transferred these phases to the use of artificial intelligence and discuss AI’s potential as a supportive tool for self-regulated learning processes, for example in goal clarification (pre-action), accompanying learning activities (action), or reflecting on one’s own learning process (post-action) (Stebner & Haverkamp, 2025). Against this background, generative artificial intelligence can be understood as a tool that facilitates active engagement and support of self-regulated learning processes.

With regard to the use of AI for learning purposes, the international state of research to date examined both positive and negative effects of artificial intelligence on student learning (Lan & Zhou, 2025). Existing studies report, among other findings, improvements in students’ learning outcomes as well as increases in engagement and motivation (Songsiengchai, 2025), which have positive effects on self-regulated learning (Li et al., 2025). In contrast, other studies point to risks, such as increased procrastination, impairments in academic performance (Abbas et al., 2024), and a weakening of critical thinking skills (Gerlich, 2025). Whether the use of generative AI, such as chatbots like ChatGPT, has positive effects on learning seems to depend on usage behavior (Güner & Er, 2025) as well as students' existing self-regulated learning skills (Han et al., 2025; Zhang et al., 2025). With regard to usage behavior, previous research has typically focus on the frequency of AI-usage, while little attention has been paid to examining learning processes from the students’ perspective. In particular, there is still very little empirical evidence available on the use of generative artificial intelligence for learning purposes among pre-service teachers. Pre-service teachers occupy a distinctive position in this context, as they currently use AI as learners themselves, while simultaneously being confronted with their future pupils’ AI-usage in their professional practice.

Building on the theoretical assumptions described earlier and the research design outlined above, this contribution analyzes pre-service teachers’ subjective descriptions of learning with AI from a qualitative perspective.

Education is increasingly benefiting from artificial intelligence (AI) in various ways to improve its quality. Yet, this situation necessitates examining its impacts on teaching and learning processes and on professional decision-making. Recent studies reveal that AI-based tools support instructional design, provide feedback, and reduce teachers’ cognitive and administrative burden (Holmes et al., 2019; OECD, 2021). Teachers most often utilise AI-based tools to structure and personalise lesson content and to design and organise activities (Kim, 2025; Heng, 2026). The literature includes numerous studies examining the use of AI-based tools in teaching; yet, fewer examine how teachers integrate AI into their pedagogy or professional learning (Chiu, Spector, & Yang, 2025). Thus, there is a gap in the literature on how AI impacts pedagogical designs, especially in the context of initial teacher training.

Lesson planning is a core competence for teachers, in which pedagogical knowledge, curriculum standards, and teaching approaches should be successfully integrated (König et al., 2020). Therefore, pre-service teachers should be well-trained to develop high-quality lesson plans that meet their students’ needs and interests. The literature suggests that AI-based tools may be a great help in improving lesson coherence, fostering differentiation, and designing objectives, activities, and assessments (Celik et al., 2023). However, empirical evidence remains scarce at this point, and there is a need for a systematic investigation to compare lesson design processes with and without AI-based tool support.

As the conference theme suggested the conditions and potentials of education research is changing, and AI integration in education is among current important changes. Therefore, the findings of the study may contribute both to the national and to the international research on the role of AI in teacher education, offering empirically grounded insights into pedagogical quality, professional learning, and technology adaptation. Austria, as the context of the study, is currently under redesigning of the curricula of education faculties, so the results of the study may significantly contribute. From the international perspective, the study aligns with wider European and global initiatives to ensure that AI integration in education is pedagogically valuable, ethically guided, and meets the needs of future teachers.

Addressing the gap in the literature, the study aims to examine pre-service teachers’ experiences regarding lesson planning, with and without the help of AcademicAI, an AI system based on OpenAI's ChatGPT. For this end, it intends to answer the following research question:

1. How do pre-service teachers experience designing lesson plans with the help of AcademicAI, in terms of effectiveness, quality, creativity, and applicability?
   1. How do they assess the lesson plans created by AcademicAI when compared to theirs?
   2. How do they evaluate their experience in terms of Performance Expectancy, Effort Expectancy, Social Influence, Facilitating Conditions, and Behavioural Intention?
   3. How do they reflect on the comparison, learning process, prompting, and long-term use of AI?

The study adopted the Unified Theory of Acceptance and Use of Technology (UTAUT; Venkatesh et al., 2003). The framework helps understand technology adoption through four subcategories: performance expectancy, effort expectancy, social influence, facilitating conditions, and behavioural intention; and it has been widely applied in educational researches (Scherer et al., 2019).

Artificial intelligence (AI) has become an integral part of many people's everyday lives and is also increasingly being used by students in the course of their studies (Amoah et al., 2025). Initial studies and meta-analyses show positive effects of AI usage on learning outcomes (Chen & Cheung, 2025) as well as on the ability to work scientifically, student engagement, and motivation (Li et al., 2025). However, these effects depend on usage behavior (Aldulaijan & Almalki, 2025). Recent international studies indicate that students’ AI use is influenced by their perceived usefulness of AI and their attitude toward AI (Kiński et al., 2025; Zhao et al., 2025). There is also initial evidence for these correlations among the group of pre-service teachers (Al Darayseh, 2023; Hoya et al., 2024). In addition, initial findings show correlations between AI usage and personality traits. Arpaci et al. (2025) found significant positive correlations with the personality dimensions of extraversion, conscientiousness, and openness, while neuroticism was associated with lower AI usage. In contrast, Kiński et al. (2025) observed that students with higher neuroticism scores used AI more frequently, so that the findings are inconsistent.

While numerous studies have already examined AI use among students in general, the group of pre-service teachers has received little attention in this context to date, particularly with regard to the connection between AI use and personality traits. Following this, the present contribution will closely examine the AI usage among pre-service teachers in Germany. This article aims to analyze correlations between AI usage and further variables such as personality traits, attitudes toward AI, behavioral intention, and perceived usefulness. The results are used to identify factors influencing AI usage among pre-service teachers and to develop approaches for teacher education. The study presented here focuses on the following research questions.

1. Which AI usage behavior do pre-service teachers exhibit and for what purposes do they use AI in the context of their studies?
2. How do attitude, behavioral intention, and perceived usefulness correlate with pre-service teachers’ AI usage?
3. How does the pre-service teachers’ AI usage correlate with various personality traits?

The theoretical basis (research question 2) is provided by Davis' (1989) Technology Acceptance Model (TAM), which has been well researched in the educational context and has also been applied to students’ use of AI (e.g., Al Darayseh, 2023). The model is based on two constructs: perceived usefulness and perceived ease of use (Davis, 1989). According to the TAM model, these two predictors influence attitude toward use, which in turn affects behavioral intention and, ultimately, actual use (Davis, 1989). Theoretically, the TAM is rooted in Fishbein and Ajzen’s (1975) Theory of Reasoned Action, which assumes that an individual’s behavior is primarily predicted by personal attitudes and subjective norms (Ajzen, 1991). Against this background, the TAM is particularly well suited to explaining students’ use of AI, as it captures key cognitive and affective evaluative processes underlying technology acceptance.

In addition, this contribution examines personality traits, since they also influence individual behavior (Digman, 1990) and there have been few studies to date on the correlations between personality traits and students' AI use (Arpaci et al., 2025). The personality traits (research question 3) are based on the Big Five concept, which distinguishes between five dimensions: conscientiousness, extraversion, agreeableness, openness, and neuroticism (Schupp & Gerlitz, 2008). In light of the previously outlined inconsistent findings, these factors are also taken into account in the present study in order to provide a more nuanced explanation of individual differences in AI use among pre-service teachers.

During a fourth-grade environmental studies lesson in a Finnish classroom, a teacher introduces a digital self-assessment platform designed to document pupils’ learning in relation to curriculum objectives. Despite careful preparation, the intended trajectory of the lesson is disrupted when the technological platform does not function as expected. The teacher faces a situation where they must find a way to overcome the moment of breakdown introduced by the malfunctioning technology.

Although such technological disruptions are common and recognisable in everyday life across European schools, research on teachers’ data literacy has predominantly conceptualised competence in relation to the intended use of educational technologies (Lee, et al. 2024). Data literacy is typically framed as a prerequisite for realising the pedagogical potential of datafication or for mitigating its risks, implicitly assuming stable technological functioning and coherent instructional trajectories (Palsa et al. in review).

Based on a ethnographic exploration of three Finnish classroom (during academic year 2024-2025), this presentation challenges these assumptions by focusing on moments when data technologies do not work as planned and by examining what teachers do pedagogically in such situations. This approach makes it possible to attend to the sociomaterial relations present in these situations (Burnett & Merchant, 2020), including relations between human actors and technological devices and platforms, as well as to the broader range of pedagogical potentials that are present, of which only some are eventually enacted.

In the vignette described above, new pedagogical potential emerges only after the teacher’s initial attempts to repair the situation by troubleshooting the technical problem in order to return to the original plan. As the technology nevertheless fails to function as intended, the breakdown opens up unanticipated pedagogical possibilities. By explicitly positioning themselves as a learner, for example, by acknowledging uncertainty, limits of knowledge, and the imperfect functioning of technology, the teacher reframes the situation as a shared learning moment. This repositioning later informs how the teacher responds to pupils’ difficulties in a subsequent task, emphasising that not-knowing is acceptable and that learning with technology involves ongoing practice rather than pre-existing mastery. What might appear, from a prescriptive data literacy perspective (Palsa et al., in review), as a failure to enact data-informed teaching instead becomes a pedagogically meaningful moment that reshapes teacher–pupil relations and reframes technology use in education as a collective and supportive practice.

This research is part of the Movement for Data Literacy (MODALITY) project funded by the Research Council of Finland.