This study aims to identify patterns of students’ interaction with Gemini AI in English speaking practice based on conversation log data. This study highlights the patterns of interaction that occur between students and Gemini AI as a medium for conversation practice, as well as their impact on students’ learning experiences. This study employed a qualitative research approach, with data collected from conversation logs between students and Gemini AI. The data were analyzed using thematic analysis to identify the main patterns emerging from student interactions with the AI. The research findings revealed seven interaction patterns that emerged: initiating a conversation, stating learning objectives, seeking language assistance, exploring topics, searching for academic information, building social relationships, and ending a conversation. In these interactions, students acted as conversation initiators, while Gemini AI served as a responsive conversation partner. Overall, this study demonstrates that Gemini AI can serve as an alternative medium for English conversation practice to support students’ fluency.
Copyrights © 2026