In Central India, land-use and land-cover (LULC) transitions have intensified over the last three decades, driven by agricultural expansion, mining activity, and peri-urban growth. This study evaluates spatio-temporal LULC change dynamics (1994-2023) and projects future scenarios (2040-2060). In this study, we used a predictor-screened Cellular Automata (CA) integrated with the Artificial Neural Network (ANN) (CA-ANN) framework. Landsat-derived LULCC maps achieved high classification reliability (κ = 0.87-0.97). Between 1994 and 2023, forest cover declined by 0.91%, primarily transitioning to agricultural and scrubland classes, while built-up and barren land expanded steadily. Model validation yielded a κ value of 0.87, indicating strong agreement between simulated and observed LULC patterns. Future projections indicate continued contraction of scrubland (−1.94%) and forest (−1.19%) by 2060, accompanied by an increase in built-up (+1.32%), agricultural (+0.58%), and barren land (+1.63%). Land degradation vulnerability index (LDVI) mapping revealed that over half of the study area falls within moderate to high vulnerability zones, particularly across forest-agriculture transition belts and mining corridors. This integrated modelling framework provides spatially explicit evidence to guide restoration prioritization, landscape planning, and monitoring efforts aligned with land degradation neutrality targets.