Understanding the dynamic risks of landslidesLandslides is imperative for land use policies in mountainous settlements. The aim of this research is to develop a framework for hazard modeling of different types of landslidesLandslides, such as those induced by natural slopes and human-modified slopes. The study, first, generates landslide susceptibility using Geospatial Artificial Intelligence (GeoAI)GeoAI techniques, such as the Random Forest algorithm. Next, it estimates the temporal probability of landslide over three different time periods using the Poisson distribution modelPoisson distribution model. Subsequently, the study integrates landslide susceptibility with temporal probability for hazard estimation. The modeling approach is applied in the Kalimpong region located in the Eastern Himalayas of IndiaIndia. Further, the study analyzes changes of buildings, roads, and agricultural lands between 2008 and 2016 under different hazard scenarios. The research identifies notable distinctions between scenarios in which landslidesLandslides are induced along natural and anthropogenic slope in similar environmental conditions. Moreover, the study exposes a concerning pattern: an increase in the construction of roads and buildings as well as a decrease in agricultural land, especially in high-hazard regions. These findings indicate the increasing risk in the study area.