Acceso usuarios
Guofang Li, Ziwen Mei
, Shiin Moon, Yilin Song, H. Henny Yeung
This study situates artificial intelligence (AI) use within broader structures of linguistic inequity in multilingual schools, where monolingual norms continue to shape learning despite increasing linguistic diversity. Challenging “AI hype” that frames educational technologies as neutral tools, the study draws on the concept of practiced language policy to advance an alignment-based account of AI efficacy across macro- (societal), meso- (programmatic), and micro-level (interactional) contexts, alongside students’ and families’ language practices. Using qualitative data from a design-based study conducted in an after-school literacy program serving elementary-aged (Grades 2–5) multilingual learners in British Columbia, Canada, the research examines students’ engagement with AI-generated multilingual literacy supports, including text-to-home-language audio resources designed to support English reading development. Findings show that the pedagogical value of AI is not inherent but contingent on the degree of alignment across nested language policy contexts. When alignment was present, AI enabled students to mobilize their home languages as both linguistic and symbolic capital, supporting English comprehension while affirming multilingual identities. When alignment was weak or disrupted by English-dominant norms, AI became a site of tension and constraint, and in some cases reproduced existing inequities. By foregrounding alignment across practiced language policy scales, this study contributes a critical framework for understanding AI in education, demonstrating that technology-mediated equity for multilingual learners depends not only on access to AI tools, but also on the ideological and policy ecologies in which such tools are embedded.
