Vocabulary acquisition is among the core elements of English as a Second Language (ESL) proficiency, and conventional teaching strategies do not always scale, offer personalized learning, or provide real-time feedback. As of late, artificial intelligence (AI) has opened new possibilities to make learners more engaged, autonomous, and receptive. The proposed research undertaking is a comparative examination of the efficacy of human-facilitated vocabulary instruction and artificial intelligence-based learning solutions in providing ESL students at different proficiency levels with learning opportunities. The investigation is conducted using a mixed-methods research design, which assesses learning outcomes, retention rates, levels of motivation, patterns of error correction, and learners’ perceptions in both instructional modes. Vocabulary gains are measured by pre- and post-tests, whereas observational data and learners’ reflections document qualitative differences in cognitive engagement and behavioral change related to learning. The paper also investigates the effects of personalization, feedback immediacy, adaptive difficulty, and multimodal content on vocabulary development under AI-based conditions. Findings reveal the capabilities of both methods that can be used in practice: human instruction provides a deeper context, emotional support, and more precise feedback, while AI tools provide a more generalizable, personalized practice by being data-driven, available all the time, and scalable. The results suggest integrating these two systems to develop a superior model of vocabulary instruction for ESL students. The study will contribute to the expanding research on the future of AI-assisted language learning, providing evidence-based insights for teachers, developers, and policymakers.
Keywords: AI-assisted learning, Comparative study, ESL vocabulary acquisition, adaptive learning systems, human-guided instruction
