The transformative power of artificial intelligence in science education: a pedagogical perspective

The rapid integration of artificial intelligence (AI) into education is reshaping science pedagogy in ways that remain insufficiently explored. While early implementations of AI in teaching show promise – particularly through intelligent tutoring systems, virtual laboratories, and adaptive learning tools – there is a critical gap in understanding how these technologies affect conceptual learning across specific science disciplines. This review identifies underexplored areas, including AI’s impact on science epistemology, equity challenges, and educators’ professional readiness. Drawing on recent literature, the paper provides discipline-specific analysis across chemistry, physics, and biology, highlighting pedagogical benefits, ethical complexities, and future research needs. To strengthen theoretical contributions, we present an original framework grounded in Vygotsky’s sociocultural learning theory and Technological Pedagogical Content Knowledge (TPACK), integrating AI as a mediational tool. This novel analytical perspective allows us to evaluate AI not just as content delivery but as a culturally situated educational artefact. Ultimately, we advocate for a human-centred, critically reflective framework that supports personalised, inquiry-based, and ethical science education.