Google rolled out an AI image-generation feature on Google Earth powered by Nano Banana 2, enabling users to create photorealistic fake scenarios overlaid on real satellite imagery. Within 48 hours, the feature was disabled after experts and users demonstrated severe misinformation risks—generating fake refugee camps, nuclear plants, and disasters on authentic maps. The incident highlights the challenge tech companies face balancing innovation with safety guardrails, particularly for tools that could weaponize trusted reference materials like Google Earth.