Search Results

ABSTRACT

This transcendental phenomenological study explores how bilingual Arab writers experience language decolonization when using artificial intelligence (AI)-generated tools in academic writing. A screening survey was distributed using snowball sampling, yielding 32 respondents. From these, nine undergraduate and postgraduate students in the United Arab Emirates were selected through criterion sampling based on their responses, willingness, and availability to participate. Drawing on interview data, the study employed cluster analysis and sentiment analysis to trace the process of decolonization through six developmental stages: initial trust in AI authority, cultural and linguistic mismatch, discomfort with standardized norms, negotiation of voice, strategic adaptation, and advocacy for inclusive AI. Participants’ narratives revealed how AI tools often reinforced dominant English norms, creating tensions around authenticity, cultural erasure, and linguistic misalignment. These tensions prompted critical reflection and selective engagement with AI feedback, fostering greater metalinguistic awareness and a reassertion of bilingual identity. The findings contribute to decolonial language theory, critical applied linguistics, and critical digital literacy by grounding how algorithmic systems influence bilingual expression. This study proposes a grounded model of AI-mediated language decolonization, with implications for culturally inclusive AI design and linguistically responsive pedagogy in multilingual academic contexts.

Keywords: language decolonization, Arab bilinguals, college students, AI generative tools, academic writing