Abstract: With the continuous expansion of oil and gas pipeline construction and the increasing complexity of operational environments, traditional risk assessment methods have gradually become limited, thus failing to satisfy the demands for accuracy, timeliness, and adaptability in modern pipeline safety management. To achieve efficient early risk warning, this paper comprehensively analyzed the development and application of data-driven risk assessment methods, summarizing the advantages, challenges, and future directions of current technological applications. The study found that data-driven models exhibit stronger adaptability and accuracy in pipeline risk assessment, allowing more detailed and real-time risk warnings. However, these methods rely highly on large volumes of high-quality data and are prone to overfitting. Moreover, the"black- box "nature of the models causes poor interpretability, bringing challenges for making critical safety-related decisions. This paper proposes several development suggestions to address the issues and challenges faced by intelligent pipeline networks in terms of both safety and economic operation.