With the increasing scale of high-voltage (HV) cable deployment and their gradual entry into the aging phase

intelligent operation and maintenance (O&M) has become an urgent requirement for ensuring the reliability of urban power grids. Conventional O&M approaches for HV cables face significant challenges

including difficulties in identifying hidden defects

tracing multi-fault coupling mechanisms

and a scarcity of effective case data. Artificial Intelligence (AI) offers a transformative opportunity to shift cable O&M from an "experience-driven" paradigm to a "data- and intelligence-driven" model

leveraging its strengths in processing complex

high-dimensional

and nonlinear datasets. This paper presents a systematic review of the current status and future prospects of AI applications in condition monitoring of HV cable systems. First

we outline the key physical parameters monitored in HV cable systems and discuss the potential of AI in this domain. Subsequently

we review the characteristics of mainstream AI techniques-particularly machine learning and deep learning-and analyze their application progress in cable condition monitoring. Furthermore

we provide an in-depth examination of representative AI applications in three core areas: partial discharge (PD) detection

evolved gas analysis (EGA) for cable outgassing

and circulating current monitoring in metallic sheaths. In response to the critical challenge of limited data availability in EGA

model optimization strategies based on data augmentation

transfer learning

and few-shot learning are discussed. Finally

we explore the technical pathways for multi-modal data fusion and intelligent diagnostic evaluation. This study demonstrates that AI

through mechanisms of knowledge self-evolution and multi-modal data integration

holds strong potential to advance HV cable systems from traditional reactive O&M toward intelligent

predictive maintenance frameworks

thereby significantly enhancing the operational reliability of entire HV cable networks.