Generic by design the missing business school perspective in AI-driven feedback research: a bibliometric analysis (2015–2025)

Research on AI-driven feedback in higher education has expanded sharply since the late-2022 release of ChatGPT, yet a foundational question remains: who is this knowledge designed for? This study examines whether a decade of scholarship has adequately served business and management education, given its distinct institutional logics, assessment genres, and professional orientations. Drawing on a Scopus-indexed corpus of 2,588 publications (2015–2025), we combine bibliometric performance analysis and science mapping with a novel disciplinary classification framework validated through dual-coder review (κ=0.87). The findings reveal a structurally imbalanced field: educational and technological perspectives account for 83.7% of knowledge production, while only 6.5% of studies are explicitly situated in business school contexts; a structural pattern that has remained stable across three technological periods, including the post-2022 generative AI surge. Interpreted through a Service-Dominant Logic lens, AI-driven feedback is predominantly treated as an operand resource for delivery efficiency rather than as an operant resource for value co-creation. As a result, management educators adopting AI-based feedback tools may inherit pedagogical assumptions not designed for their disciplinary context. This study contributes a reproducible framework for diagnosing disciplinary imbalances in AI-in-education research and develops a targeted agenda that repositions business education as a primary site of inquiry.