<![CDATA[This study aims to identify, classify, and systematically analyze various problems in the utilization of Artificial Intelligence (AI) in Islamic education in Indonesia, as well as to formulate preventive and remedial solutions. The study employs a qualitative approach with a combined design, using the Systematic Literature Review (SLR) method based on PRISMA guidelines, integrated with a narrative-conceptual literature study. Data were obtained from 14 field research articles indexed in Scholar, Sinta, and Scopus within the 2020–2026 period, and were analyzed using content analysis techniques and issue trees. The findings indicate that the primary problems lie in two domains: cognitive-affective psychological impacts and violations of academic ethics and privacy. In the cognitive-affective aspect, the study identifies symptoms such as bias in religious understanding, a decline in critical thinking skills, a tendency to accept AI-generated outputs dogmatically, and the weakening of pedagogical-spiritual relationships between teachers and students. In the ethical domain, issues include AI-assisted plagiarism and the misuse of deepfake technology in religious content. The contributing factors include the epistemic limitations of AI systems, the lack of comprehensive regulation, academic pressure, and increasing user dependency. This study recommends an epistemological reconstruction based on maqasid al-shari‘ah, positioning AI as a supporting instrument (wasilah), supported by pentahelix synergy and the implementation of the ADDIE instructional design model, so that AI integration can be carried out ethically, accountably, and in alignment with the objectives of Islamic education.]]>
