Ethical and technical challenges of AI-driven geospatial applications in civil, environmental, and geomatics engineering

The field of civil, environmental, and geomatics engineering (CEGE) is changing as a result of the combination of artificial intelligence (AI) and geospatial technologies, or GeoAI. Advanced predictive modeling, spatial automation, and data-driven infrastructure planning at previously unheard-of scales are made possible by this convergence. However, despite these advancements, there are still a number of technical difficulties and unexamined ethical issues with using AI in geospatial contexts. In order to assess how GeoAI is operationalized within CEGE and how ethical and technical aspects intersect throughout its development lifecycle, this study provides a critical synthesis of 30 peer-reviewed publications (2013–2025). From conceptual ambiguity in defining GeoAI to application trends in hydrological modeling, urban planning, and environmental surveillance, five main themes emerge: persistent technical limitations like data heterogeneity, poor model transferability, and high computational demands; ethical risks like algorithmic bias, surveillance-driven privacy erosion, model opacity, and accountability gaps; and the lack of strong governance frameworks, especially in underrepresented global regions. The review shows a disjointed body of knowledge where engineering optimization and ethical foresight are often separated. In response, this study presents an ethics-by-design-based research agenda with a focus on sustainability-centered evaluation metrics, contextualized model development, and participatory co-design. To incorporate ethical protections into the GeoAI pipeline, a conceptual roadmap is put forth. A Critical GeoAI paradigm one that promotes social justice, openness, and ecological responsibility in built environment systems in addition to technical sophistication is advocated in the study's conclusion.