Landslide forecasting, the prediction of spatio-temporal probability of slope instability, is essential for effective Landslide Early warning systems. Landslides are a significant threat globally, yet data scarcity limits traditional forecasting methods. This review paper examines global approaches for landslide forecasting using Earth Observation (EO) data, which provides comprehensive spatial coverage and enhanced data accessibility. We review the use of satellite-derived precipitation products (using radar, passive microwave, and infrared) for rainfall threshold determination, the use of radar and optical imagery for Soil moisture estimation, DEM generation, and other variables integrating into landslide susceptibility maps for ML/ANN based forecasting models, We also review GNSS and InSAR data for ground displacement prediction, aiding in landslide forecasting, while infrared thermal data to detect landslide precursors. We conclude by highlighting current limitations and proposing future directions of research.