<p>Artificial intelligence (AI), particularly machine learning and deep learning, has emerged as a critical tool of food safety management systems (FSMSs). In recent years, AI has been increasingly applied to food production, processing, packaging, storage, and distribution, demonstrating clear advantages over conventional reactive strategies. However, most applications remain confined to process-specific implementations. Moreover, differences in data characteristics, combined with limited interoperability and fragmented AI deployment, prevent analytical outputs from being systematically linked, thereby restricting the effective management of cascading food safety risks. This review critically synthesizes recent advances in AI-driven food safety management across the food supply chain and systematically explores the structural barriers that hinder their integration into a coherent FSMSs framework. By reframing AI not as a collection of isolated analytical tools but as an integrated decision support infrastructure, this review proposes strategic directions for interoperable AI deployment that connect data flows and analytical models.</p>