The AI Act and the future of STEM education in Europe: rethinking pedagogy, assessment, and teacher agency

Generative artificial intelligence (AI) is swiftly transforming STEM education, presenting pedagogical, ethical, and regulatory challenges. The Artificial Intelligence Act, a comprehensive legislative framework emphasizing transparency, accountability, and human oversight, is implemented throughout Europe. This paper analyzes the necessity of re-evaluating STEM education in light of the convergence between generative AI and the AI Act. The paper posits that, grounded in pedagogy, evaluation, and teacher agency, AI functions as an epistemic actor that actively participates in the generation and validation of knowledge. In AI-dominant environments, conventional instructional and evaluation frameworks are inadequate and necessitate new process-oriented, authentic, and ethical methodologies. It also underscores the increasing importance of teachers as mediators of technological and regulatory forces. The interplay among policy, practice, and professional agency is illustrated using a conceptual framework. The paper concludes with implications for educators, teacher training, and policy, contributing to the discourse on European STEM education.