‘AI-Learner Partnership’: Psychological mechanisms and developmental trajectories of EFL learners’ writing agency in GenAI-assisted courses

Abstract This study investigates how EFL learners construct, negotiate, and develop writing agency when engaging with generative artificial intelligence (GenAI) tools in second language writing courses, conceptualizing these technologies as cultural mediational means within an integrated sociocultural and positive psychological theoretical framework. Eleven Chinese EFL university learners were observed over a 16-week period as they completed multiple academic writing tasks using GenAI assistance. Data collection employed a qualitative multi-method approach including screen recordings of writing sessions, stimulated recall interviews, reflective journals, and sequential drafts with AI interaction logs. Learners’ agency trajectories evolved through three developmental phases: initial exploration (characterized by novelty, dependency, and emergent self-efficacy), strategic adaptation (marked by selective tool use and growing writing-specific psychological capital), and deliberate appropriation (reflecting personalized integration and flourishing techno-authorial identity). Four patterns of agency manifestation emerged: transitional agency (shifting dependencies with emerging metacognitive awareness and increased resilience), distributed agency (strategic negotiation of textual authority fostering growth mindsets), reflective agency (critical evaluation of AI contributions coupled with self-determination), and transformative agency (reconceptualization of writer identities within human-AI collaboration leading to authentic engagement and eudaimonic well-being). Cognitive processes, affective dimensions, positive psychological resources, and task design characteristics significantly mediated how learners constructed agency through interactions with technological affordances. This study extends theoretical understanding of agency in GenAI-assisted educational contexts by illuminating the dynamic, developmental nature of writing agency and identifying specific patterns of agency manifestation in human-AI collaborative writing, while bridging sociocultural perspectives with positive psychological constructs that support optimal functioning.