A pulsating heat pipe (PHP) is a passive two phase heat transfer device without any mechanical parts and additional power inputs. Due to its simple structure, PHP has a wide range of applications such as electronics, satellite systems, solar technology, cryogenic and nuclear fields, etc. However, the partial understanding of its complex operating mechanism limits their widespread utilization in the industry. Researchers have developed many mathematical models but there is still much that remains unknown. To address this limitation, machine learning is an alternative tool for the modeling of PHPs and many researchers have been working on it for several years. This review explores the application of machine learning techniques across the different aspects of operational aspects of PHPs. Key discussion includes the comparative effectiveness of different machine learning models and the suitability of these models on extrapolating findings across various experimental setups. Here, further, we will discuss the advantages of machine learning as well as their potential for advancing PHP technology.