Acute kidney injury (AKI) is a common, complex, and potentially modifiable complication in hospitals. The incidence is around 5% in the wards and approximately 35% in the ICU setting. The mortality ranges from 18% to 80% in different diseases or operations. Hospital stays are prolonged in those survivors, and healthcare costs are hugely expanded. More and more older adults become acute kidney disease and then chronic kidney disease, resulting in a fragile status or morality. One report suggested poor AKI care and delays in diagnosis or even failures to recognize AKI. Electronic alert systems have been applied in many hospitals to improve the early recognition of AKI. And the care bundles or consultations following the alert might provide the convenience of patient care. However, the result showed heterogenesis in different hospitals. Moreover, predicting AKI and avoiding AKI are what physicians want to do next. Artificial intelligence is going viral, and the application of machine learning is also increasing in hospitals. The daily care might be changed after the intervention of real-time AKI prediction.