Language Production Environments under Digital Constraint: Adaptive Repertoires and Ethical AI Boundary-Making in a Mobile-First Multilingual and Inclusive Learners Context

Abstract

This qualitative study examines how diverse undergraduate learners navigate systemic infrastructural barriers—such as poor connectivity and device inequality—to achieve equitable language production. Framed by Universal Design for Learning (UDL) and Self-Regulated Learning (SRL), the research explores adaptive practices in a mobile-first, multilingual higher education context. Data from semi-structured interviews and reflective journals with 30 English language students reveal that marginalized learners actively reorganize their production environments. Findings indicate students strategically adapt or abandon inaccessible digital workflows to preserve core academic standards. Furthermore, learners utilize multilingual repertoires and generative AI as critical, stage-specific assistive scaffolding during planning phases, while fiercely protecting their authorship in final outputs. This bounded AI engagement demonstrates a sophisticated inclusive digital literacy. Ultimately, the study argues that educational equity requires institutions to validate flexible, asynchronous workflows and support responsibility-driven AI engagement over rigid technological compliance.