The rapid emergence of generative AI (GenAI) tools, including Large Language Models (LLMs), is reshaping higher education and challenging traditional approaches to language teaching. College language instructors are increasingly required to integrate AI into their pedagogical practice, yet successful implementation depends critically on their own AI literacy. This study conceptualizes AI literacy as encompassing technical, pedagogical, critical, and ethical competencies, which collectively enable teachers to design, evaluate, and implement AI-enhanced learning experiences. Using a case study of an English debate contest involving 90 first-year undergraduates, this research examines how teachers’ varying levels of AI literacy influence instructional design, student engagement with AI tools, and learning outcomes. Findings suggest that higher AI literacy facilitates transformative teaching practices, characterized by personalized, collaborative, and authentic learning experiences, while lower literacy constrains effective AI integration. The study underscores the importance of professional development programs that cultivate comprehensive AI literacy, enabling language educators to harness AI responsibly and effectively, ultimately fostering enhanced student learning and critical thinking in the AI-driven educational landscape.