The Impact of AI-Assisted Writing Tools on the Quality of English for Specific Purposes Writing and Learner Strategies: An Empirical Study of Blended Learning

This study aims to explore the application effects of AI-assisted writing tools in English for Specific Purposes (ESP) writing instruction and their impact on learners' writing strategies. Using a blended learning empirical research design, this study recruited 128 students majoring in economics and management at a college. Through a 16-week teaching experiment, by combining writing tests, questionnaires, and in-depth interviews, this study systematically explored the specific mechanisms by which AI-assisted tools affect the quality of ESP writing. The results show that AI-assisted writing tools have the potential to improve learners' writing performance in five dimensions: grammatical accuracy, lexical richness, register appropriateness, content completeness, and structural logic. Simultaneously, these tools encourage learners to develop three adaptive writing strategies: human-computer collaboration, process monitoring, and metacognitive regulation. Interestingly, there were significant intergroup differences in tool use and strategy selection based on learners' English proficiency levels. This study provides empirical evidence and practical insights for AI-assisted ESP writing instruction in blended learning environments.