Generative AI tools, such as ChatGPT, have rapidly influenced diverse fields, including human geography. They offer opportunities for advanced computational analyses, qualitative data handling, and the generation of synthetic data for sensitive topics. Their capabilities, such as text-to-image modeling, can potentially democratize quantitative methods and assist in auditing built environments, reducing the need for extensive coding or high-performance computing infrastructure. However, generative AI tools also pose challenges. Hallucination effects—where AI confidently fabricates misinformation—can mislead users, while political, social, and geographic biases risk uneven outcomes and interpretations. Geographic biases manifest as reduced accuracy or specificity for certain areas, particularly rural or low-population-density regions. Similarly, environmental justice data queries for less populous U.S. counties reveal ChatGPT’s limited local knowledge. Although methods like retrieval-augmented generation may address some issues, the root lies in disparities in training data quality and coverage. Human geographers must carefully weigh the potential of generative AI against these risks, adopt strategies for responsible use, and emphasize improved, equitable data collection practices to ensure positive and reliable impacts on human geography research.