Recent breakthroughs in generative AI have created an enormous realm of possibilities and opportunities for developing systems and solutions in language teaching and learning. Generative AI is providing a vast domain of resources for educators and language teachers, enabling them to transform the process of teaching and learning, doing what was once impossible in traditional approaches. The aim of this study was to investigate and compare the effectiveness of three methodological and procedural approaches to content selection in developing university students’ reading comprehension skills: using AI-generated texts, using standard internationally renowned textbooks and adoption or adaptation of freely available texts by teachers. To this end, of 39 BA students enrolling in an EAP course focusing on reading skill at Ilam University, Iran, 33 students were placed in three homogeneous classes of 11, organized after a placement test on general proficiency. In one class, AI generated texts made by teacher’s prompts to AI chatbots ChatGPT and Google Gemini were used as the main material, while in one other class Select Readings Intermediate textbook, a reading coursebook published by Oxford University Press was used, and in another class, the teacher adopted and adapted texts from freely available online resources. Class timing was identical in three classes. Upon completion of the course, students took an extensive reading comprehension test developed based on the syllabus common between the classes. The test results showed that students in the class using AI-generated texts slightly outperformed the class using the textbook, and students in both AI-assisted and textbook-based classes demonstrated better performance compared to the students in the class with adapted and adopted materials. Further studies in different contexts and settings, with higher numbers of learners, and investigating other skills, subskills or components, as well as courses and programs with other aims, objectives and syllabi are recommended.
