This study examines artificial intelligence ethics from an Islamic perspective, focusing on the distinction between machine and human responsibility. Using a descriptive-analytical method, supported by comparison, it relates contemporary international principles to the Islamic concepts of moral accountability, trust, justice, non-harm, and the objectives of Sharia. The study concludes that current AI systems lack consciousness, intention, and moral capacity; therefore, they cannot bear full ethical or religious responsibility. Responsibility remains human and should be distributed among designers, developers, institutions, operators, and users according to their knowledge, foreseeability, and control. Responsible AI consequently requires continuous human oversight, data protection, non-discrimination, risk assessment, explainability, accessible objection mechanisms, and compensation for harm. These safeguards ensure that AI serves human welfare, protects dignity, and advances the public interest without becoming a means of avoiding accountability.................. Keywords:............... Artificial intelligence, ethical responsibility, Islamic thought, Maqasid al-Sharia, human oversight.