Re-orienting science education in the generative AI era: From inquiry to practice

Generative Artificial Intelligence (GenAI) is reshaping scientific research at both epistemological and methodological levels. However, the phenomenon of GenAI hallucination presents critical challenges to the credibility and validity of GenAI-enhanced scientific research. In response, science education must evolve from open-ended inquiry towards evidence-based practice, cultivating students’ ability to critically conduct scientific research with GenAI. Therefore, the paper proposes a three-tiered pedagogical framework that integrates: (1) Integrative STEM education as the theoretical foundation, (2) Maker-centred learning and Socioscientific Issues (SSI) as core content, and (3) Project-based learning as the implementation model. The paper further contends that realising this framework requires systemic efforts in three key areas: empirical validation, teacher development, and policy formulation, to ensure that science education in the GenAI era is effective, scalable, equitable, and sustainable.