The growing integration of artificial intelligence (AI) in education worldwide has raised concerns about the potential de-professionalization and de-skilling of teachers, particularly as core pedagogical responsibilities, such as lesson planning, knowledge transmission, and assessment, are increasingly supported or automated by AI-enhanced systems. As these tools begin to generate instructional materials and recommendations at scale, pedagogical control may shift away from teachers, positioning them less as professional decision-makers and more as implementers of algorithmically generated content (Cukurova, 2025). This risk underscores the importance of teacher agency and professional judgment in AI-mediated classrooms.

In this context, a growing body of literature argues that educators require a specific set of AI-related competencies to effectively navigate AI integration in ways that protect their professional role and support pedagogically meaningful use (Ng et al., 2023; Cukurova & Miao, 2024; Cukurova, 2025). Even though there are competency frameworks available for teachers (e.g., Cukurova & Miao, 2024), most existing studies and professional development (PD) initiatives have focused on enhancing teachers’ AI competency without adopting a comprehensive framework (e.g., Vazhayil et al., 2019; Ding et al., 2024). Moreover, many PD programs emphasize the technical aspects of AI and ethical concerns, while giving comparatively less attention to the pedagogical integration of AI into teaching and learning.

To address these gaps, the present study aims to develop preservice teachers’ AI competency by adopting an agency-oriented approach. The AI Competency Framework for Teachers proposed by UNESCO (Cukurova & Miao, 2024) served as the primary guide for the design of the AI competency learning module for preservice teachers. The UNESCO Framework employs a competency-based approach to support teachers in integrating AI into teaching with a human-centered approach and is structured as a two-dimensional matrix comprising five competency aspects (AI foundations, human-centered mindset, AI ethics, AI applications, and AI pedagogy) across three progression levels (Acquire, Deepen, Create). To conceptualize teacher agency, the study adopts Priestley et al.’s (2015) ecological model of teacher agency. The model describes teacher agency as a situated achievement that emerges through the interaction of three dimensions: the iterational dimension (past experiences, beliefs, and knowledge), the practical-evaluative dimension (judgements under present constraints and affordances), and the projective dimension (orientations toward future goals).

Accordingly, the study investigates how preservice teachers’ AI competencies are developed and understood through the ecological model of teacher agency by implementing an agency-promoting AI competency learning module informed by the iterational, projective, and practical-evaluative dimensions of teacher agency. Regarding this purpose, the research questions are:

RQ1: How does participation in the agency-promoting AI competency module (the AICompA Module) improve preservice teachers’ AI competency?

RQ2: How does preservice teachers’ agency develop through the ecological model of agency (iterational, practical-evaluative, projective) in the context of the AICompA Module?

A qualitative case study design will be used to investigate how preservice teachers (PSTs) develop AI competency through the lens of the ecological model of teacher agency. Participants will be the students (PSTs) enrolled in an undergraduate educational technology course, which aims to provide pedagogical foundations and practical competencies for integrating technology into classroom practice. An agency-promoting AI competency learning module (AICompA) will be embedded in the course to foster PSTs’ AI competencies for professional AI use. The module's structure was guided by the UNESCO AI Competency Framework for Teachers (AI CFT) and comprises five interconnected learning units- AI foundations, a human-centered mindset of AI, the ethics of AI, AI applications, and AI pedagogy. Over an eight-week implementation, PSTs will engage in hands-on activities including scenario-based analyses, development of ethical use guidelines, and AI-supported instructional design tasks. The learning module was also designed to support teachers’ agency by surfacing their experiences and beliefs related to AI, engaging them with contextual situations that require making judgements under constraints, and supporting their professional roles as designers of future AI-enhanced learning environments. Multiple data sources will be collected to understand the development of AI competency and agency improvement. Before the module, participants will complete the Teacher AI Competence Self-Efficacy Scale (TAICS) adopted by Chiu et al. (2025) to determine their initial AI competency. During the module, participant-produced artifacts will be collected as supporting qualitative data. At the end of each unit, participants will respond to a set of reflective questions designed to understand what they learned, how they justify AI integration-related decisions in educational contexts, and how they position themselves as future teachers making decisions with/around AI (agency-related reflections). After the module, TAICS will be re-administered to measure changes in self-reported AI competency. Finally, semi-structured interviews will be conducted to examine participants’ development of teacher agency concerning their AI competencies. Data will be analyzed using both quantitative and qualitative analysis. For RQ1, TAICS pre-post scores will be compared to assess change in AI competency, while reflections, artifacts, and interviews will be analyzed through thematic analysis to understand competency development over time and to explain which competencies were strengthened. For RQ2, reflections and interviews will be analyzed using a hybrid coding: deductive coding guided by Priestley et al.’s ecological model of teacher agency (iterational, practical-evaluative, projective) and inductive coding to identify emergent themes which demonstrate how PSTs’ agency evolves in relation to AI competency.

The study is expected to provide three interrelated sets of outcomes concerning (a) preservice teachers’ AI competency development, (b) shifts in teacher agency conceptualized ecologically, and (c) design implications for preservice teacher education. Firstly, in relation to AI competency (RQ1), participation in the AI competency learning module is expected to produce measurable gains across AI competency aspects, including AI foundations, human-centered mindset, AI ethics, AI applications, and AI pedagogy. Beyond increases in AI knowledge (e.g., understanding AI concepts, limitations, and ethical issues), development is anticipated in practice-oriented competencies, such as the ability to evaluate AI outputs critically, make pedagogically justified choices, and design learning activities that integrate AI. Regarding teacher agency (RQ2), the module is expected to strengthen preservice teachers’ agency across the iterational, practical-evaluative, and projective dimensions. In terms of iterational teacher agency, preservice teachers may report increased confidence and a more articulated belief system about responsible AI use. With respect to practical-evaluative agency, they are expected to demonstrate improved capacity to identify constraints (e.g., bias, data privacy, institutional expectations) and to justify situated decisions about when and how AI should be used. In the projective dimension, preservice teachers are expected to articulate clearer professional roles by specifying how they intend to position themselves as teachers in relation to AI systems. This may include envisioning themselves not merely as users of AI tools, but as pedagogical designers who set educational goals, select appropriate uses of AI, and take responsibility for instructional and assessment decisions. Lastly, the study aims to provide evidence-based design principles for AI competency instruction in preservice teacher education that promote agency and align with established frameworks. Practically, it seeks to offer insights for teacher educators and policymakers who are looking to integrate AI into teacher education in a human-centered manner.

Cukurova, M. (2025). Promoting and Protecting Teacher Agency in the Age of Artificial Intelligence. Cukurova, M., & Miao, F. (2024). AI competency framework for teachers. UNESCO Publishing. Ding, A. C. E., Shi, L., Yang, H., & Choi, I. (2024). Enhancing teacher AI literacy and integration through different types of cases in teacher professional development. Computers and Education Open, 6, 100178. Ng, D. T. K., Leung, J. K. L., Su, J., Ng, R. C. W., & Chu, S. K. W. (2023). Teachers’ AI digital competencies and twenty-first-century skills in the post-pandemic world. Educational Technology Research and Development, 71(1), 137-161. Priestley, M. R., Biesta, G., & Robinson, S. (2015). Teacher agency: An ecological approach. Bloomsbury Publishing. Vazhayil, A., Shetty, R., Bhavani, R. R., & Akshay, N. (2019, December). Focusing on teacher education to introduce AI in schools: Perspectives and illustrative findings. In 2019 IEEE tenth international conference on Technology for Education (T4E) (pp. 71-77).