The use of AI in historical studies has shown an impressive trajectory in recent years, thanks to breakneck progress in computing power, data availability, and machine learning algorithms. The current qualitative study will explore the main problems of using AI tools, e.g. NLP and ML in studying Islamic historical documents. Via an exhaustive literature review and an application of different NLP and ML techniques to a self- selected corpus of relevant sources, the author is going to state the difficulties of AI use in historical studies, such as the accuracy and difficulty of classical Arabic language, availability of historical digital texts, and the need for considerable domain knowledge in the process of reading the historical results. Then enumerate the findings for the pursuits of the AI investigations in Islamic history such as Linguistic challenges, Data sparsity and lack of digitized of historical sources, and the Need for human expertise in the cultural context and propose strategic solutions including collaborative interdisciplinary work, advance digitization efforts, provision of specific AI tools to the historical domain, and ML models development, in addition to the real need to the collaboration between the human abilities and the AI analytic tools, to get a better result.