The increasing expansion of artificial intelligence and its growing reliance on automated processing of personal data have created new challenges in the field of civil liability. Algorithmic decision-making, especially in the absence of effective human oversight, can lead to breaches of privacy, material and moral damages, and the weakening of individual rights; a matter that poses serious difficulties for traditional rules of civil liability in the field of attributing harmful acts and proving causation. Accordingly, the main question of the article is to what extent the legal systems of Iran and Afghanistan are able to organize the civil liability arising from the automated processing of personal data by artificial intelligence by relying on the existing rules, and what model is more appropriate for compensation. The aim of the research is a comparative analysis of the approach of Iranian and Afghan law to civil liability arising from the processing of personal data and an evaluation of the capacity of international documents to complete and guide these systems. The research method is descriptive-analytical with a comparative approach and based on the study of domestic legal sources and international documents. The research findings show that although the general rules of civil liability in Iran and Afghanistan are applicable to damages caused by artificial intelligence, in practice they face limitations, and the tendency towards risk-based liability and the principle of accountability can be more efficient. The result of the research indicates that aligning with international standards and redefining civil liability will play an effective role in protecting personal data and effectively compensating damages.