The increase in popularity of Generative Artificial Intelligence Chatbots, or GACs, has created a potentially fruitful opportunity to enhance teaching English as a Foreign Language (EFL). This study investigated the possibility of using GACs to give EFL students metalinguistic guidance (MG) in linguistics courses. Language competency gaps, a lack of individualized engagement, and low metacognitive abilities are common challenges EFL students face in linguistics courses. Feedback has been suggested as a potential solution to these issues in previous studies; nevertheless, conventional corrective feedback (CF) might not fully satisfy the demands of EFL students. In order to address these obstacles, the current study suggested a metalinguistic guiding (MG)-based GAC approach. Using a quasi-experimental approach with pretest and posttest setups, this study evaluated the learning achievement, reflective performance, perception, and metacognitive awareness of EFL students exposed to either CF-based GAC or MG-based GAC. According to the study's findings, the MG-based GAC group performed better than the CF-based GAC group in terms of learning achievement, reflective performance, and perceptual and metacognitive awareness. The GAC's immediate educational usefulness and potential as a pedagogical tool for shaping cognitive processes are highlighted by its successful application in helping EFL students gain metacognitive awareness. This study contributes significantly to the growing body of knowledge about the use of GAC in educational settings by providing empirical evidence of the effectiveness of GAC in terms of delivering MG to EFL students.