Assessing the use of AI tools for EFL exam preparation at Saudi universities: efficiency, benefits, and challenges

This study examines EFL teachers’ adoption of AI tools in exam preparation practices in Saudi universities. Employing the TPACK and UTAUT models, the research explores AI applications, perceived advantages, and encountered challenges. Sixty-one EFL teachers from two universities in Saudi Arabia participated in this study via an online survey. Premised on two relevant theoretical frameworks to examine EFL teachers’ use of AI in exam writing, the scope of the study was limited to the efficiency, benefits, and challenges educators may experience during exam preparation. The findings ascertained that AI tools such as ChatGPT, Quizbot, Quizlet, ExamSoft, GEMINI, and Gemini AI were commonly used for exam preparation. The findings ascertained that ChatGPT emerged as a popular instrument for creating questions showing alignment with TPACK through blended technological, pedagogical, and content knowledge. The findings also revealed that AI use centered on lower-order questioning like multiple-choice/true-false, offering efficiency gains. Greater efficiency and more significant time-savings were also reported. Emerged challenges were effort expectancy resulting from customization obstacles to meeting curriculum standards, prompting inaccuracies, lack of institutional training, and limitations in alignment with learning outcomes. Therefore, findings underscore balanced reliance on technologies and expertise while ensuring assessments align with learning outcomes. The study also highlighted intrinsic limitations, including the necessity for substantial human oversight over complicated tasks such as essay prompt design, accuracy concerns regarding generated output, and ethical implications involving bias/transparency worries. Recommendations involve strengthened teacher preparation, AI capacity improvements, and ethical guidelines development to steer AI integration judiciously. This contributes meaningfully to evolving research investigating AI roles within evolving evaluative processes, particularly under Vision 2030 in Saudi Arabia.