Conceptualizing Writing in the Digital Age: A Systematic Review of Research on AI-mediated EFL Analytical Writing


  •  Lidan Chen    

Abstract

This systematic review examines how writing is conceptualized in studies on AI-mediated English as a Foreign Language (EFL) analytical writing and how these conceptualizations are enacted in pedagogical practice. Drawing on Ivanič’s (2004) Discourses of Writing and Engeström’s (1987) Activity Theory, 20 peer-reviewed studies published between 2021 and 2024 across four main academic databases are systematically analyzed. The review examines dominant writing discourses, patterns of AI use across pre-, during, and post-writing stages, and the components of the activity system that mediate these practices. Findings show both stability and change in AI-mediated writing research. Skills and process discourses remain predominant, with 8 studies each. These studies frame writing as linguistic accuracy and procedural mastery, supported by automated evaluation tools. Meanwhile, discourses of creativity, social practices, and sociopolitical perspectives are emerging, which suggests a shift towards dialogic, collaborative, and ethically aware conceptions of writing. AI use was most common in the post-writing stage, with 18 studies, in which tools serve either as revision assistants or automated assessors. These patterns reveal tensions between accuracy and agency, and efficiency and authorship. From an Activity Theory perspective, AI acts as a dual mediating artifact that stabilizes traditional instructional routines while reconfiguring feedback structures, classroom roles, and learner agency. The review concludes that AI-mediated writing should not be seen solely as a technological enhancement but as a site for cultivating critical awareness, reflective engagement, and ethical authorship in EFL writing pedagogy.



This work is licensed under a Creative Commons Attribution 4.0 License.
  • ISSN(Print): 1923-869X
  • ISSN(Online): 1923-8703
  • Started: 2011
  • Frequency: bimonthly

Journal Metrics

Google-based Impact Factor (2021): 1.43

h-index (July 2022): 45

i10-index (July 2022): 283

h5-index (2017-2021): 25

h5-median (2017-2021): 37

Learn more

Contact