Main Article Content
Abstract (in English)
Artificial intelligence (AI) is increasingly being investigated as a tool for improving poultry production, health monitoring, environmental management, and farm automation. This narrative review synthesizes scientific evidence on AI applications in poultry production and evaluates their potential applicability, infrastructure requirements, limitations, and future opportunities in Afghanistan. Literature was searched using Google Scholar, Scopus, PubMed, Web of Science, IranMedex, and Magiran using terms related to artificial intelligence, machine learning, poultry production, disease detection, computer vision, Internet of Things, robotics, animal welfare, and Afghanistan. The initial search identified 66 potentially relevant records, and 30 publications were included in the narrative synthesis after screening and eligibility assessment. Current evidence indicates that AI can support early disease and abnormal-behavior detection, environmental monitoring, precision management of feed and production, and robotic farm operations. However, successful implementation depends on data quality, reliable electricity and connectivity, suitable sensors and computing infrastructure, technical expertise, and economic feasibility. Importantly, the available literature does not provide sufficient evidence to conclude that AI is already widely adopted in Afghanistan’s commercial poultry sector. Therefore, AI adoption in Afghanistan should be approached through locally validated, gradual, and economically appropriate applications, beginning with digital record keeping, environmental sensing, computer-vision monitoring, and decision-support systems. Future studies should evaluate these technologies under real Afghan poultry-farming conditions.