Skip to content

Generative AI in University EFL Writing: Tensions among Efficiency, Language Proficiency, and Learner Agency

Sep 2026 · Journal of Linguistics & Cultural Studies · 0 citations · 10 references

TL;DR

It is argued that the central issue is not whether students use AI, but how responsibility for thinking and decision-making is distributed between students and AI, and proposes a conceptual model in which the depth of AI involvement interacts with learner agency to shape learning outcomes.

Abstract

The rapid diffusion of generative artificial intelligence (GenAI) is reshaping university-level English writing by lowering the linguistic and cognitive costs of text production. Large language models can generate and revise prose, suggest vocabulary, explain grammatical problems, reorganize paragraphs, and provide immediate feedback. These affordances can improve writing efficiency and the surface quality of students' texts, particularly when learners face linguistic constraints or limited access to individualized feedback. Yet the educational value of GenAI cannot be reduced to technological empowerment. When AI performs ideation, language production, organization, and revision on behalf of learners, opportunities for language output and cognitive engagement may be reduced. This paper examines the resulting tension among writing efficiency, language proficiency, and learner agency through the lenses of second language acquisition, cognitive load, constructivist learning, and self-regulated learning. It argues that the central issue is not whether students use AI, but how responsibility for thinking and decision-making is distributed between students and AI. The paper proposes a conceptual model in which the depth of AI involvement interacts with learner agency to shape learning outcomes. The most productive condition is a collaborative development zone in which students retain ownership of goals, arguments, and final judgments while using AI as a linguistic and cognitive scaffold. The paper further argues for process-based assessment, critical AI literacy, staged AI access, and a redefinition of EFL writing competence. In an AI-mediated environment, effective writing competence extends beyond language generation to include evaluation, strategic collaboration, and meaning reconstruction.

Read PDF

Similar papers

Open access Sep 2026

AI-mediated writing and durable learning: a psychological theory of calibration, bypassing and uptake

Artificial intelligence (AI)-mediated writing provides a critical case for educational psychology because it reveals how learners regulate cognition, effort, confidence, motivation, and beliefs about competence when fluent external support is available. In English as a foreign language (EFL) writing, such support can c...

Zihang Lan, Chen Niu · 1 citation
Open access Aug 2026

Generative AI Mediation in EFL Oral and Written Production: Colombian Student and Teacher Perception

How students and teachers in a Colombian university language program perceived GenAI-supported production was explored and four patterns emerged: GenAI as a rehearsal and revision partner; tension between polished output and language ownership; teacher mediation shifting toward critical language awareness; and the need...

Dionelio Jesus Moreno Villalobos · 0 citations
Review Open access Aug 2026

Learner cognition and behavioral engagement in GenAI-mediated professional language education: a critical integrative review

It is argued that the central educational challenge is not whether GenAI improves single-task language performance, but how learners develop calibrated trust, critical judgment, self-regulated feedback use, and professional agency in human–AI language-learning environments.

Xiong Wang · 0 citations

Related blog posts

Google DeepMind Blog Nov 18, 2026

Genie 4 is now available

We are rolling out a new Project Genie experience to eligible Google AI Ultra subscribers, opening access to our new Genie 4 world model.

Microsoft Research Blog Nov 16, 2026

TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient

AI coding agents are increasingly help developers fix bugs and build features largely on their own, but every interaction with their models consumes tokens. In each session, an agent sends instructions, tool descriptions, and session history to its model. Most efforts to reduce this cost focus on individual tasks. Yet developers return to the same […] The post TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.