The use of Generative AI (GenAI) in Computer-Assisted Language Learning (CALL) can provide new ways of generating content, but it will require a robust pedagogical evaluation. Using a mixed-methods approach, this study examines the views of N = 15 English language teachers and N = 35 university students on AI-generated grammar materials; the study is based on the TAM, which provided the basis for an examination of three dimensions: Curriculum Quality, Student Engagement and Technology Acceptance. Quantitatively, results indicate that 71% of students prefer AI-based methods to traditionally delivered instruction due to a high degree of perceived interactivity (M = 3.53). While teachers agreed that there was a positive impact on student engagement (M = 3.67), they expressed a moderate level of comfort with the use of AI (M = 3.47), and were concerned about both the lack of cultural context of the AI-generated materials, and the lack of accuracy of the grammar presented in these materials. Overall, the study found a cautious optimism that supports the need for a blended model of instruction that utilizes AI as an engaging tool, under teacher-pedagogical supervision.