Codification of scholarly values in academic publishing: AI governance and institutional tensions

Abstract

This article examines the current artificial intelligence (AI) policies by five prominent academic publishers—Elsevier, Springer Nature, Taylor & Francis, SAGE Publishing, and Cambridge University Press— interpreting them as efforts to codify values as generative AI becomes embedded in academic communication. Through comparative qualitative analysis, the study identifies core values consistently operationalized across these policies, including exclusive human authorship, mandatory disclosure, non-delegable accountability, and confidentiality in peer review. As technological change disrupts traditional norms, these policies act as both regulatory tools and ethical anchors, stabilizing the identity and legitimacy of academic institutions. However, this codification creates structural tensions. Turning ethical principles into formal rules produces unintended consequences, from adding new layers of scholarly labor to reinforcing existing institutional inequalities. In addressing both epistemic risks and the tensions that arise from governance responses, the article advocates for AI governance that is reflexive, inclusive, and context-sensitive. Such models should preserve institutional integrity while respecting the evolving partnership between humans and machines in the production of knowledge.