Although geographic information systems (GIS) are widely used in tsunami studies, they continue to present several limitations, requiring further work. Due to the complexity of the data and the need for accurate real-time analysis, the process of both effectively assessing tsunami risk and hazard mapping can be challenging. The ability of machine learning (ML) to automate data processing and identify patterns in large and complex datasets will significantly improve the accuracy of real-time prediction models in the future for relevant and comprehensive tsunami risk analysis. However, to date, researchers have done little to explore the application of learning models within GIS. This article presents the studies that used GIS and ML algorithms in tsunami research. Then, sheds light on the possible application of ML algorithms in combination with GIS as an emerging solution to overcome the current deficiencies of GIS in their tsunami analyses. Consequently, by merging GIS visualization capabilities with the predictive use of ML algorithm performance, not only is the accuracy of tsunami prediction models improved, but also the efficiency of the decision-making process for hotspot identification. These benefits will help authorities to better understand and manage tsunami risks, strengthening the resilience of populations against these disasters.