Presentation From Blank Prompts to Pedagogical Control: Supporting CI-Aligned Lesson Planning with AI more

As generative AI tools become increasingly accessible, many language teachers report a common frustration: although AI can generate lesson materials quickly, the output often fails to align with pedagogical intent, learner readiness, or classroom realities. This practice-oriented presentation argues that the issue lies not in AI capability, but in a mismatch between how teachers plan instruction and how AI systems are typically prompted. Grounded in established principles of Comprehensible Input (CI) and the Zone of Proximal Development (ZPD), the session foregrounds teacher cognition rather than technology. It examines how experienced instructors routinely make intuitive decisions about learner readiness—deciding when students are ready to understand, use, or extend language—and how time pressure makes it difficult to translate these judgments into explicit lesson plans. Through classroom examples and a brief demonstration, the presentation illustrates how AI can function as a supervised planning assistant when teachers provide pedagogically meaningful constraints. Emphasis is placed on revision, selection, and rejection of AI output as essential professional practices. A short, optional planning routine demonstrates how AI can reduce planning friction while preserving instructional philosophy. The session contributes to CALL practice by modeling how teachers can integrate AI into existing workflows without relinquishing pedagogical control.