Artificial Intelligence (AI) in the Field of Tahfiz: A Simulation Study of ChatGPT Interaction in Qur’anic Murājaʿah

Abstract

The Qur’anic revision (murājaʿah) process is a crucial element in [1]strengthening memorization, yet its implementation is often hindered by limited time for tasmiʿ sessions with teachers or peers, a lack of suitable teaching aids, and repetitive review approaches. The advancement of artificial intelligence (AI) technologies presents new opportunities to support self-directed learning in the field of Qur’anic memorization (tahfiz). This study evaluates the potential and limitations of using ChatGPT—a large language model (LLM) developed by OpenAI—as a supportive tool in the murajaah process. A qualitative approach was employed through interactive simulations involving two sets of random cloze-style questions spanning all 30 juzʾ of the Qur’an. The data were analyzed based on question structure and style, distribution of verses by juzʾ, difficulty level, and the textual accuracy of Qur’anic content. The findings indicate that ChatGPT is capable of generating a variety of verse continuation questions but tends to favor shorter verses from certain juzʾ. Some errors were also identified, such as incorrect recitation markers, invalid verses, and inaccurate juzʾ references. These issues highlight the need for supplementary guidance, such as digital mushaf input and more precise prompts. Although ChatGPT cannot replace the role of a teacher, it shows potential as an interactive tool for self-guided murājaʿah, particularly for students outside formal learning environments. This study lays the foundation for developing human-AI interaction guidelines and AI-based teaching materials in tahfiz education that align with the principles of Islamic education.