Personalized Language Learning with AI and Large Language Models: Tailoring Instruction to Learner Prerequisites

Personalization has long been a central goal in language education, reflecting the wide variation learners bring in prior linguistic knowledge, cognitive resources, and learning goals. Recent advances in artificial intelligence (AI), particularly large language models (LLMs), have intensified interest in technology-supported personalized language learning by enabling adaptive interaction, feedback, and content generation. However, prevailing approaches to personalization in AI-supported systems are frequently driven by surface-level performance indicators and interactional fluency, offering limited insight into learners’ instructional readiness. This paper introduces a prerequisite-centered framework that reconceptualizes personalization as a problem of instructional alignment grounded in learner readiness rather than adaptive responsiveness alone. Drawing on second language acquisition (SLA) and language pedagogy, the framework conceptualizes learner prerequisites as encompassing linguistic, cognitive-affective, and contextual dimensions that mediate instructional effectiveness. Using this framework as a normative evaluative lens, the paper critically examines how different AI-supported approaches—including LLM-based tools—conceptualize and enact instructional tailoring. The analysis demonstrates that while contemporary AI systems provide substantial flexibility and interactional richness, they often lack principled mechanisms for aligning instructional decisions with learners’ developmental preparedness. The paper concludes by outlining pedagogical implications for prerequisite-aware personalization and theoretically grounded directions for aligning AI-supported language learning with sustained language development, instructional coherence, and teacher agency.