Multiple co-occurring drivers of global change are poised to cause global biodiversity loss and the deterioration of ecosystem services and resilience. In both agricultural and wild ecosystems, bees supply essential pollination services. However, the conservation of bees faces challenges due to the scarcity of occurrence data, which is often biased spatially and taxonomically. Despite bee monitoring initiatives gaining public and political support in recent years, they lag current data needs for facilitating the inclusion of bee diversity in contemporary biodiversity conservation decisions. By leveraging existing collection records and community science data through single-, joint, and stacked species distribution models, I predicted current and future bee distributions and species richness. I found that bee species’ responses to global change vary based on their life history traits and environmental preferences. In the desert southwest, a global hotspot for bee diversity, the relative threats of urbanization, utility-scale solar energy development and climate change vary across an elevational gradient. Despite extensive sampling in this region, many bee species, especially those at the greatest risk of decline, lack sufficient data to be included in models. I show that distribution models offer a valuable tool to address these data limitations. For example, using a specialist oil bee, Macropis nuda, I exhibit the utility of species distribution models for informing targeted collection of rare species. In both temperate North America and the Desert Southwest, my models predict that bee distributions are likely to shift to track their climate niches, although the extent of predicted redistribution is variable. I demonstrate that with careful data cleaning, correction for spatial bias, and models robust to presence-only data with small sample sizes, it is feasible to generate estimates of bee distributions and diversity using currently available data. Continued refinement of modeling methods for data-limited species and additional data collection are needed to ensure ecologically important pollinators are included in contemporary biodiversity conservation initiatives and land use decisions, including renewable energy development.