<![CDATA[Generative artificial intelligence (AI) poses challenges for Islamic universities seeking to reconcile innovation, academic integrity, and religious responsibility. This study examines how Islamic ethical principles are interpreted and translated into generative AI governance at two Indonesian state Islamic universities, and how the findings qualify an existing multilevel Living Islam model. A qualitative multiple-case design draws on 37 semi-structured interviews, 18 institutional documents, and two co-design workshops conducted between February and June 2026 at UIN Sunan Kalijaga Yogyakarta and UIN Maulana Malik Ibrahim Malang. Four empirical themes concern AI's productive yet subordinate role, a gap between declared values and enforceable procedures, lecturers' frontline governance work, and tensions among efficiency, integrity, privacy, and equitable access. These patterns inform a six-stage analytic cycle linking normative interpretation, stakeholder deliberation, institutional translation, pedagogical enactment, ethical accountability, and reflective evaluation. The proposed cycle is only partially evidenced in both cases and should not be confused with a fully implemented policy system. The study conceptualizes Living Islamic Ethics as an institutionally mediated process and distinguishes general AI safeguards from their specifically Islamic moral justification. Its findings are limited by a two-case design, reliance on reported practice, and incomplete documentation of spontaneous ethical references and workshop model development.]]>
