The integration of generative AI like ChatGPT into EFL pedagogy presents both opportunities for fostering higher-order thinking skills and risks to academic integrity. A research gap exists regarding the simultaneous impact of ChatGPT on the crucial triad of critical thinking, creativity, and self-reflection within the Indonesian EFL context. This study aimed to fill this gap by quantitatively measuring the effect of ChatGPT on these skills and qualitatively exploring students' perceptions of the learning process. The study used a mixed-methods sequential explanatory design. Participants were 100 undergraduate students, randomly assigned to either an experimental group (n=50) or a control group (n=50). Data were collected using three validated instruments: the Critical Thinking Scale, the Creative Thinking Scale, and the Reflective Thinking Scale. In addition, a semi-structured interview guide was used to obtain qualitative data. Quantitative data were analyzed using ANCOVA, and qualitative data using thematic analysis. The findings revealed that the ChatGPT group achieved statistically significant gains in critical, creative, and reflective thinking scores compared to the control group. Qualitative results revealed a duality in student perceptions; they valued the AI for fostering skills through idea generation and safe practice, but expressed concerns about risks such as cognitive offloading and skill atrophy.
