VietTESOL International Convention 2026

Compulsory English courses in Vietnamese technical universities face a structural tension that has proven difficult to resolve through conventional means. Engineering undergraduates, for whom English occupies an obligatory rather than vocational role in the curriculum, tend to disengage from assessment processes that offer delayed, uniform feedback with little responsiveness to individual learning trajectories. This paper reports on a semester-long institutional pilot that introduced an AI-powered adaptive learning platform into a foundational English program at a technical university in northern Vietnam, with the aim of examining whether adaptive technology could meaningfully alter that pattern.
The platform under study adjusts content difficulty and task sequencing in response to real-time learner performance data and delivers formative feedback automatically upon task completion, independent of instructor input. The pilot was conducted across four classes during Semester 2 of the 2025-2026 academic year, involving 135 third-year engineering undergraduates and eight instructors. A mixed-methods design was employed, drawing on platform usage analytics, semi-structured teacher interviews, and student perception surveys.
The findings suggest two substantive shifts. Students demonstrated greater willingness to reattempt tasks and engage with corrective feedback when it was delivered automatically, a pattern attributed in part to the reduced evaluative pressure of system-generated responses. Instructors reported a meaningful reallocation of professional time away from routine error correction toward communicative and task-based instruction. The paper concludes by examining the institutional conditions required to sustain these shifts beyond a pilot phase, with implications for technology integration policy in technical higher education.