Evaluating Generative AI in Indonesian Education: Pragmatist and Social Reconstructionist Perspectives on Pedagogical and Structural Implications

Abstract

Despite the rapid integration of Generative Artificial Intelligence (GenAI) in education, existing scholarship remains predominantly focused on technical performance, with limited attention to its normative and philosophical implications, particularly in non-Western contexts. This study addresses this gap by applying a dual philosophical framework grounded in Deweyan pragmatism and Countsian social reconstructionism to evaluate GenAI-mediated pedagogy. Using a qualitative conceptual design, this study employs a systematic thematic synthesis of peer-reviewed literature to examine Generative Artificial Intelligence (GenAI) through the philosophical lenses of pragmatism and social reconstructionism. The findings reveal three key insights. First, GenAI demonstrates educational value only when pedagogically designed to augment, rather than replace, active student inquiry. Second, without deliberate policy intervention, its implementation risks amplifying existing digital inequalities in Indonesia. Third, the absence of a coherent national AI governance framework exposes educational systems to risks related to algorithmic bias and data injustice. Based on these findings, this study proposes a three-dimensional evaluative model integrating pedagogical quality, equity, and governance, offering a structured framework for guiding responsible GenAI adoption in developing educational systems.