Generative artificial intelligence (GenAI) is increasingly applied across geographic research, yet existing studies remain fragmented and lack a field-level synthesis. This study addresses this gap through a bibliometric analysis of 891 peer-reviewed journal articles published between 2022 and 2025 and indexed in Scopus. The results indicate a rapid expansion of GenAI-related geographical research following the diffusion of large language models, alongside strong spatial concentration in a small number of countries and institutions. Five major research themes are identified, including LLM-based GeoAI and GIS workflows, Earth observation and remote sensing, knowledge-driven urban analytics, educational and scholarly communication contexts, and tool-oriented geospatial platforms. Temporal patterns suggest a shift from early exploratory studies toward workflow-level integration. Highly cited contributions concentrate in education, remote sensing, and GIScience, while intellectual foundations draw on both foundational AI architectures and long-standing geographic measurement and modeling. By providing a systematic and large-scale mapping of the structure and evolution of GenAI in geography, this study extends beyond existing narrative and application-focused reviews and offers an integrated account of the field’s development. It also identifies key research gaps for future research.
Ng et al. (Wed,) studied this question.