Sepsis associated-acute kidney injury (SA-AKI) is a common complication of sepsis, which has a high incidence and is closely related to a poor prognosis. However, delayed diagnosis and non-specific treatments make it difficult to systematically manage SA-AKI. Based on massive clinical data, machine learning could build prediction models, which provide alarms and suggestions for the clinical decision support system. Although there are still many challenges such as poor interpretability, it has shown clinical application value in SA-SKI risk prediction, imaging diagnosis, subtype identification, prognosis assessment, and so on. Based on a brief introduction of machine learning, this article reviews the application, limitations, and future directions of machine learning in the diagnosis and treatment of SA-AKI, and explores the possibility of machine learning in the medical field, in order to promote the development of precision medicine and intelligent medicine.