Structured review uncovers five GeoAI cartographic dysfunctions in automated mapmaking, highlighting an operational literacy framework for human-AI collaboration.
Key Points
To establish an integrative typology of cartographic dysfunctions caused by geospatial artificial intelligence and provide an operational literacy framework to identify and remediate them.
Conducted a structured review of cartographic, GIScience, and AI-ethics literature published from 2020 through 2025.
Applied the conceptual framework to an empirical case study analyzing the Tacoma/Seattle/Beijing CycleGAN model.
Conducted an initial round of expert consultation to achieve preliminary validation of the proposed literacy matrix.
Classified GeoAI cartographic dysfunctions into two operational axes: artifact-visible errors (geographic bias, cartographic hallucination, deepfake geography) and production-hidden flaws (model collapse, infrastructural opacity).
Mapped dysfunctions against four core cartographic literacy domains—semiological, geodetic, datalogical, and ethical—supported by four curricular instruments for remediation.
Redefined cartographic agency as distributed across the human-AI production network to account for upstream corpus assemblers, annotators, and model developers.
Cite This Study
Hayat Yagouni-Bouzar (2026) studied this question.