Google removed a new Google Earth image-editing feature one day after launch because it could turn trusted satellite views into convincing AI-altered scenes. The tool used Nano Banana 2 to generate changes on top of real Earth imagery from text prompts.
Researchers quickly showed why that was risky. Examples included refugees near the Mexican border, a bomb crater near a hospital in Gaza, and other fabricated scenes placed on recognizable geography. Google said generated images were watermarked and did not appear in the main Earth experience for other users.
The problem is that Google Earth is often used by journalists, investigators, and researchers as a reference for verifying real-world images. If altered screenshots spread outside the product, a watermark may not be enough to preserve trust.
Google said it is rolling back the feature while it works on stronger guardrails. The episode shows how adding generative tools to trusted reference products can create risks that are different from ordinary image generation.