This study investigated the use of AI to generate compelling input for EFL learners, focusing on interest and comprehensibility. College EFL learners (n=26) prompted AI models to generate 260 stories based on frequency-based vocabulary lists. These AI-generated stories served as reading input, evaluated for their suitability as extensive reading materials. Students read, summarized, and rated the stories for interest, reporting their use of level adjustment prompts (LAP) to evaluate whether the stories matched their proficiency level. Students also provided their motivation levels to determine if story ratings were influenced by their general engagement with language learning or the content quality of the stories. Level adjustment prompts were used for 160 of the 260 stories (62%), with higher use among lower-proficiency learners (A1: 85%; C1: 0%). Interest ratings averaged 3.65 (SD=0.91, between “average” and “a little interesting”), with no significant variation by motivation level (p>0.05). The study demonstrates a preliminary approach to generating level-specific stories based on frequency-based vocabulary and highlights the potential of AI-generated stories as a low-cost, adaptable resource for language learning programs, while emphasizing the importance of pedagogical strategies and ongoing refinement of AI prompts to enhance level-appropriateness.
George Loetter (2026) studied this question.