<![CDATA[This study examines the effectiveness of integrating Project-Based Learning (PBL) with Gligglish AI in improving the English-speaking skills of second-semester students in the Information Systems Research Program at a private university in Indonesia. In this context, many students experience persistent challenges in speaking development, including linguistic difficulties such as limited vocabulary, grammatical control, and pronunciation accuracy, as well as psychological barriers such as low self-confidence and fear of making mistakes. The study employed a mixed-methods sequential explanatory design. Quantitative data were collected through pre- and post-test assessments using a validated speaking performance rubric. In contrast, qualitative data were obtained through semi-structured interviews to explore students’ learning experiences and perceptions. The instructional intervention involved implementing Project-Based Learning activities supported by Gligglish AI, which facilitated interactive speaking practice, pronunciation feedback, and opportunities for autonomous learning. The participants were two intact classes (n = 52), divided into an experimental and a control group. The quantitative findings indicate substantial improvements in students’ speaking performance following the intervention, with fluency increasing by 31.4%, pronunciation by 28.6%, grammar by 24.8%, and vocabulary by 26.7%. The most pronounced improvement was observed in overall speaking performance, which increased by 47.8% from the pre-test to the post-test. Qualitative findings further reveal that students perceived the integration of PBL and Gligglish AI as engaging, relevant to their academic field, and effective in enhancing confidence, motivation, and active participation. These findings suggest that integrating AI-assisted speaking tools within a project-based learning framework offers a promising instructional approach for improving English-speaking skills in private higher education contexts.]]>
