Google pauses Earth AI image tool over misinformation risks

Google has rolled back a new Google Earth feature that allowed users to create AI-generated images on top of real satellite, aerial or 3D imagery. The company paused the tool less than 48 hours after announcing the integration of its Nano Banana 2 image generator, saying it was working on stronger guardrails after seeing shared screenshots that appeared to violate its policies.

The feature let users enter text prompts to create fabricated scenarios tied to real-world map coordinates, then download and share the results. Tests produced images including a collapsed Eiffel Tower, a sinkhole at the Great Pyramid of Giza and Russian tanks in Ukraine’s capital. Misinformation experts warned that false images could inherit the perceived legitimacy of Google Earth, especially in conflict zones or fast-moving crises where reliable information is limited.

Google said images made with its tools include invisible watermarks and can be checked through Gemini or Lens. However, tests found those checks could sometimes be bypassed, while external AI detection tools did not always identify the images as AI-generated. Specialists warned that weakening trust in satellite imagery could be especially damaging because it has long been viewed as a reliable source for understanding hard-to-reach areas.