Abstract
The integration of Artificial Intelligence (AI) into English as a Second Language (ESL) assessments has revolutionized traditional practices by offering efficiency, accuracy, and personalized learning pathways. This study employs a mixed-methods approach to evaluate the effectiveness of AI tools, such as Grammarly, Duolingo, and Write & Improve, in improving ESL learners' proficiency across writing, reading, speaking, and listening skills. Quantitative findings from 150 learners show significant improvements in writing (16.6%) and reading (13.8%), while gains in speaking (5.4%) and listening (4.2%) remain modest, reflecting the limitations of AI in handling nuanced oral communication. Qualitative insights from 20 instructors reveal challenges, including algorithmic bias, cultural insensitivity, and concerns over data privacy. Despite these issues, AI tools are praised for reducing grading time and providing instant feedback. The study emphasizes the need for ethical guidelines, equitable access, and human oversight to address existing limitations and ensure inclusive educational outcomes. Additionally, it highlights the digital divide, where socio-economic disparities limit access to premium AI tools, exacerbating educational inequalities. By combining quantitative data with qualitative insights, this research provides a comprehensive understanding of AI's role in ESL education. It advocates for a balanced integration of AI, positioning it as a complementary tool that amplifies human expertise rather than replacing it. This study contributes to ongoing discussions on the ethical and practical implications of AI in education, offering recommendations for policymakers, educators, and developers to optimize its potential.