Google has paused its recently introduced AI-powered image generation feature in Google Earth less than two days after its global release. The decision follows reports of users sharing AI-generated images that appeared to breach Google's policies.
The feature utilized Google's Nano Banana 2 image-generation model, enabling users to create AI-enhanced scenes based on satellite, aerial, and 3D imagery within Google Earth. Users could zoom into any location, select "Create image," and describe the desired scene.
Reason for Suspension
Google stated it temporarily disabled the tool to strengthen safeguards after seeing some generated images that violated its standards. The company emphasized that these AI-generated images were never integrated into the public Google Earth platform and were clearly watermarked as AI creations.
"We know that people uniquely trust Google Earth for a reliable view of the world," Google said. "While many used this feature for constructive purposes, some shared misleading images that undermined trust. We are rolling back the feature as we implement stronger guardrails."
Potential Applications and Risks
At launch, Google highlighted the tool's potential to assist in education, architecture, urban planning, and historical visualizations by combining real-world imagery with AI-generated content. However, concerns quickly emerged over the misuse of the feature to fabricate convincing but false geographic scenes.
Investigations by researchers and journalists demonstrated how the tool could produce realistic yet fictional images, such as a collapsed Eiffel Tower or a sinkhole at the Great Pyramid of Giza. Experts warned that while watermarked within Google Earth, screenshots shared externally could spread misinformation by detaching context.
Geospatial analysts cautioned that linking AI-generated images to real locations risks eroding public confidence in satellite imagery, a trusted resource for journalists, humanitarian groups, and researchers.
Next Steps
Google has not provided a timeline for the feature's return but confirmed it will remain offline until enhanced protections are in place to prevent misuse while preserving its legitimate educational and professional benefits.
