<![CDATA[Listening comprehension remains a critical yet persistently underdeveloped skill in English as a Foreign Language (EFL) higher education, particularly in Indonesian universities where authentic materials, native speaker input, and personalized learning pathways are severely limited. Despite growing interest in AI-assisted language learning, few studies have translated proof-of-concept AI integrations into scalable, pedagogically grounded mobile applications. This study aimed to develop, validate, and implement ListenToFy, a mobile EdTech application that integrates ChatGPT for adaptive content generation and VoxBox Text-to-Speech for authentic, native speaker audio, to transform Advanced Listening instruction. Employing a Research and Development (R&D) design using the ADDIE model, the study was conducted across two Muhammadiyah universities in Indonesia involving 34 EFL students and two lecturers. Expert validation yielded an overall score of 85.7%, indicating high validity. Usability testing produced a System Usability Scale score of 82.1, classified as Excellent. Pre- and post-test results demonstrated statistically significant listening comprehension gains (mean gain: +14.0 points; p < .001; Cohen’s d = 0.79–0.84), and qualitative interviews identified four key themes: authentic audio experience, personalized learning pathways, instructor empowerment, and technology integration challenges. ListenToFy represents a replicable, dual-AI EdTech model that meaningfully advances listening instruction reform in Indonesian and comparable Southeast Asian EFL higher education contexts.]]>
