BEYOND GRAMMAR CORRECTION:GENERATIVE AI AND THE CULTIVATION OF HIGHER-ORDER THINKING IN EFL WRITING

The rapid integration of generative artificial intelligence (GenAI) into educational settings has raised new possibilities and challenges for second language acquisition, particularly in the domain of academic writing. While most existing applications emphasize lower-order concerns such as grammar correction, spelling, and vocabulary enhancement, less is known about how GenAI can be harnessed to cultivate higher-order thinking skills in English as a Foreign Language (EFL) writing contexts. This study addresses that gap by examining the extent to which GenAI tools can support analysis, evaluation, and creation with core dimensions of higher-order cognition as outlined in the Revised Bloom’s Taxonomy. Guided by Cognitive Learning Theory and Sociocultural perspectives on mediated learning, the study employed a mixed-methods, quasi-experimental design with 80 intermediate-level EFL university students. Participants were randomly assigned to an experimental group (n = 40), which engaged in GenAI-assisted writing and revision tasks, and a control group (n = 40), which received traditional instructor feedback with a focus on grammar and mechanics. Over the course of twelve weeks, students produced weekly essays, engaged in iterative revisions, and, for the experimental group, critically interacted with AI-generated feedback targeting content development, argument structure, and rhetorical clarity. Data sources included pre- and post-intervention writing assessments evaluated against a rubric emphasizing higher-order components (argumentation, coherence, and originality), weekly reflective journals, and semi-structured interviews with a subset of participants. Quantitative results revealed statistically significant gains in the experimental group’s higher-order writing dimensions compared to the control group, with notable improvements in argument development, logical coherence, and the generation of original insights. Qualitative findings further highlighted that students using GenAI engaged in deeper revision cycles, demonstrated increased metacognitive awareness, and reported heightened confidence in idea generation and critical evaluation of their own writing. Nevertheless, some risks were identified, including over-reliance on AI suggestions, occasional uncritical acceptance of machine-generated text, and uncertainties surrounding academic integrity. These findings underscore the dual role of GenAI as both a scaffold and a potential crutch in EFL writing pedagogy. The study concludes that when thoughtfully integrated into instructional design, GenAI can extend beyond its remedial function of grammar correction to become a catalyst for higher-order thinking. Pedagogical implications include the need for explicit training on critical engagement with AI outputs, structured reflective activities to reinforce metacognitive skills, and assessment frameworks that reward originality and critical reasoning rather than surface-level accuracy alone. By shifting the focus from linguistic correctness to cognitive depth, educators can leverage GenAI not merely as an editing tool but as an active partner in fostering analytical, evaluative, and creative writing competencies among EFL learners.