Google has temporarily disabled a newly introduced AI-powered image generation feature for Google Earth images just a day after launching it, following concerns over the creation of highly realistic fake images.
The tool was driven by Google’s Nano Banana 2 AI model and enables users to build realistic images from text prompts using Google Earth’s satellite, aerial, and 3D map data.
Google built the tool for legitimate uses such as urban planning, education, architecture, and design visualization. But it wasn’t long before users realized they could generate convincing images of imaginary disasters and conflict zones, including bombed cities, floods, refugee camps, and war scenes.
According to Google, many professionals used the feature as intended, but the company also identified AI-generated content that appeared to breach its usage policies.
Although every AI-generated image carried a visible watermark and was never integrated into the public fake Google Earth images platform, screenshots of the fabricated scenes could still be shared across social media, increasing the risk of misinformation.
Google has taken the feature offline while it builds stronger safeguards to prevent possible misuse. The company aims to reintroduce the tool once it can better ensure responsible and policy-compliant use.
