AI policy as de facto language policy: towards AI policy literacy for the language education sector

Abstract

Abstract This article argues that institutional AI policy functions as de facto language policy when it regulates how language is used, taught, and assessed, making AI policy literacy an essential professional capacity for the language education sector. As AI governance increasingly shapes the practices and future of language education, educators need to move beyond interpreting and complying with institutional policies to critically examining how these policies are framed, enacted, governed, and revised, and actively participating in their development. Yet institutional AI policymaking has been predominantly framed through technical, ethical, and academic integrity concerns, leaving language expertise underrepresented and the sector’s policy capacity largely unexamined. Drawing on an expert panel of senior language education leaders, this article develops a conceptual framework of AI policy literacy comprising six interconnected dimensions: textual, discursive, enactment, governance, equity, and adaptive literacies. It argues that language educators have a legitimate claim to professional jurisdiction in AI governance, but that exercising this jurisdiction requires deliberate and systematic policy capacity-building. Rather than positioning language educators merely as stakeholders affected by institutional decisions, the framework provides a foundation for their informed participation and leadership in AI policymaking. By conceptualising AI policy literacy as both a professional repertoire and a means of strengthening sectoral agency, the article positions AI policy research as an emerging area of inquiry that brings language expertise into institutional AI governance. Its implications extend beyond language education to other disciplinary and professional communities seeking greater agency in shaping the policies that govern their practices.

The authors' abstract, as published at the source. Language Policy, 2026 · DOI ↗