AI Listening in Low-Resource Classrooms: Offline Tools and Their Effectiveness

In the context of global efforts to integrate artificial intelligence (AI) into language learning, low-resource classrooms—particularly in rural or underserved regions—face unique challenges due to limited internet connectivity, insufficient hardware, and minimal teacher training. This study explores the potential of offline or minimally connected AI tools to support listening comprehension among Uzbek EFL learners in such environments. Conducted across four rural schools, the research involved 90 secondary students and compared outcomes between a control group using traditional textbook audio and an experimental group utilizing offline AI-supported listening tools, including speech-enabled mobile apps, downloadable AI-generated dialogues, and USB-based practice materials.

Findings revealed that students using offline AI tools demonstrated significantly higher engagement, motivation, and listening accuracy compared to the control group. The availability of features such as interactive playback, real-time feedback (even without internet), and adaptable content played a crucial role in boosting comprehension and learner autonomy. Teacher interviews and classroom observations indicated that even with limited digital literacy, educators successfully integrated these tools and witnessed improved student participation and confidence. However, barriers such as the lack of localized content and insufficient training remained. The study concludes that low-tech AI integration offers an effective, scalable, and equitable pathway to enhance listening instruction in under-resourced contexts, provided that it is supported by teacher capacity-building and contextualized content design.