Between gatekeeping and coping: Iranian EFL students’ lived experience of AI-mediated writing assessment

Abstract

The rapid proliferation of Generative Artificial Intelligence (GenAI) has destabilized foundational assumptions in second language writing assessment. The dominant scholarly response—concentrating on detection, psychometric validity, and pedagogical integration—has systematically overlooked the subjective, affective, and moral experience of learners themselves. This study addresses that gap by investigating the lived experience of Iranian EFL university students navigating high-stakes academic writing assessment in a context where GenAI tools are widely available yet institutionally unregulated. Employing Interpretive Phenomenological Analysis (IPA) guided by person-centered theory and Self-Determination Theory, semi-structured interviews were conducted in Persian with purposively selected undergraduate students at a public Iranian university. Findings revealed three superordinate themes: a structural double-bind wherein the assessment system simultaneously demands flawless independent writing and incentivizes the use of tools it condemns; a fractured self-experience marked by diminished authorship, the isolating weight of secrecy, and an oscillation between relief from writing anxiety and shame at felt inauthenticity; and the construction of personal moral geographies through which students navigate ethical ambiguity in the absence of institutional guidance. The findings demonstrate that current assessment arrangements generate systemic incongruence and thwart the basic psychological needs for autonomy, competence, and relatedness. Drawing on these findings, the study proposes three principles for human-centered assessment design—transparency, integration, and relationality—as a contextually grounded contribution to ongoing discussions about reconciling the demands of accountability with the imperative to support learner well-being in GenAI-mediated assessment environments.