This study examines the potential of artificial intelligence (AI) in delivering personalized feedback to lower secondary students in Hanoi. Using a mixed-methods approach that combines surveys and case study analysis, data were collected from 60 students and seven teachers at two public schools. Results show that the AI platform significantly reduced grading time (by 45%) and improved student satisfaction with writing feedback (by 62%). The system provided tailored responses to each student’s strengths and weaknesses in grammar, coherence, and vocabulary. Furthermore, in advanced mathematics classes, AI helped identify recurring error patterns to support group-based remediation strategies. Despite several challenges related to system compatibility and teacher readiness, AI holds great promise in Vietnamese education when integrated with proper infrastructure and pedagogical adaptation.