<p>Landslides pose a major threat to the Himalayan mountains, and their frequency is expected to rise under climate change. This study assesses present-day and climate change–driven landslide susceptibility in Kinnaur district, Himachal Pradesh, India, by integrating geomorphological and environmental conditioning factors with CMIP6 precipitation projections. A database of 12 landslide conditioning factor maps was prepared, including land use/land cover, elevation, slope, aspect, curvature, drainage density, geology, topographic wetness index (TWI), geomorphology, NDVI, soil texture, and rainfall. A Random Forest approach was used to generate the susceptibility map and was validated using Accuracy, Precision, Recall, MCC, Kappa, and AUC. The model performed strongly (Accuracy: 91.78%; AUC: 0.97), indicating reliable prediction in complex Himalayan terrain. The baseline map shows that 85.66% of the district falls in very low susceptibility, while ~ 11% lies in high to very high susceptibility, concentrated along steep valley side-slopes, river corridors, sensitive lithologies, and road alignments. Future trends were evaluated using precipitation projections from three CMIP6 models (EC-Earth3, MPI-ESM1-2&#xa0;h, and NorESM2-MM) under SSP245 and SSP585 for near (2021–2045), mid (2046–2070), and far (2071–2095) periods. Results indicate increasing seasonal precipitation, especially during monsoon and post-monsoon, with a corresponding expansion of highly and very highly susceptible zones, most pronounced under SSP585. Under far-future SSP585, very high susceptibility may exceed 11% of the district, indicating a worsening risk scenario.</p> Graphical Abstract <p></p> <p>This graphical abstract presents a visual overview of the methodology and key outcomes of the study on climate-induced landslide susceptibility in the Kinnaur district of Himachal Pradesh, India. The graphical framework integrates geospatial datasets, machine learning techniques, and future climate projections to assess current and potential future landslide risks in a complex Himalayan terrain. The first component of the graphical abstract highlights the study area and the data sources used, including landslide inventory information, digital elevation models, satellite-derived environmental factors, and precipitation datasets. The second component illustrates the analytical workflow in which twelve landslide conditioning factors such as slope, elevation, geology, rainfall, land use/land cover, NDVI, drainage density, soil texture, geomorphology, curvature, aspect, and topographic wetness index are processed within a GIS environment. The central element of the graphical abstract depicts the Random Forest machine learning model used for landslide susceptibility mapping. This model analyses the nonlinear relationships between landslide occurrences and environmental conditioning factors to generate reliable susceptibility predictions. The results section visually represents the spatial distribution of susceptibility classes ranging from very low to very high across the Kinnaur district. Finally, the graphical abstract highlights the integration of CMIP6 precipitation projections under SSP245 and SSP585 scenarios to evaluate potential future changes in landslide susceptibility. Overall, the graphical abstract provides a concise and visually engaging summary of the research workflow and its findings, emphasizing the increasing landslide susceptibility associated with projected climate-driven precipitation changes in the Himalayan region.</p>