The RAIL Framework: Reflection as Epistemic Formation in the Education of GenAI-Enhanced Science Teachers; implications for assessment. In Artificial Intelligence and Innovation in Education: Ethical and Technological Dimensions, AI-Education 2025 Official Conference Proceedings https://doi.org/10.5281/zenodo.20738242

Description

The RAIL Framework: Reflection as Epistemic Formation in the Education of GenAI-Enhanced Science Teachers; implications for assessment

 

Kallia Katsampoxaki-Hodgetts, University of Ioannina

[email protected]

in Artificial Intelligence and Innovation in Education: Ethical and Technological Dimensions

AI-Education 2025 Official Conference Proceedings

 

 

Abstract. GenAI is beginning to reshape teacher education beyond technical skill or

efficiency, touching instead on how novice educators interpret their work and the

responsibilities attached to it. This study examines how pre-service science teachers

conceptualise such shifting expectations while designing AI enhanced lessons.

Working within the traditions of Learning by Design and the pedagogy of

multiliteracies, and informed by recent proposals for AI-extended TPACK, we analysed

fifteen reflective accounts written after science student-teachers engaged in iterative

prompting, evaluation, and redesign. The reflections reveal a profession in formation:

students used AI to explore representational possibilities and test instructional ideas,

yet they also encountered moments of hesitation, critique, and selective resistance.

Across the narratives, reflection emerges not as a procedural add-on but as an epistemic

activity through which participants negotiated the authority of AI, articulated the

rationale behind their pedagogical choices, and considered the ethical implications of

adopting algorithmic suggestions in science classrooms. Drawing on these insights, the

paper advances the Reflective-AI Literacies (RAIL) framework, which foregrounds

epistemic reflexivity, multiliteracies awareness, and responsible AI agency as

interlinked dimensions of contemporary science-teacher professionalism. Embedded

reflection also highlights the value of process-based assessment, making visible the

evaluative work through which student-teachers shape their designs.

Keywords. AI Pedagogical content knowledge, Epistemic reflexivity, assessment,

lesson design, ethical deliberations, AI literacy framework

Authors

DOI: 10.5281/zenodo.20738243

Publication Date: 2026-06-17

Back to publications list


About