Google has decided to disable a new artificial intelligence-based tool in its mapping service Google Earth. The feature was removed less than a day after its launch due to sharp criticism from security researchers, as users began to massively generate realistic fake satellite images of disasters, war zones and secret sites.
Google has decided to disable a new artificial intelligence-based tool in its mapping service Google Earth. The feature was removed less than a day after its launch due to sharp criticism from security researchers, as users began to massively generate realistic fake satellite images of disasters, war zones and secret sites.
This is reported by the Financial Times.
The new feature involved integrating the Nano Banana generative model directly into the Google Earth interface. The developers envisioned that users would be able to virtually reimagine any place on Earth using simple text prompts.
Today we’re bringing AI image generation with Nano Banana to Google Earth, letting you virtually reimagine anywhere in the real world. 🍌 https://t.co/zsgvAnbYrl
— Google Earth (@googleearth) July 30, 2026
For the first time, you can generate custom images using Google Earth’s satellite, aerial, and 3D imagery. 🌎 pic.twitter.com/yed45vLZOo
However, immediately after its release, OSINT experts and users demonstrated how the tool was used to create disinformation. Social media began to spread fabricated satellite images of a crater from an explosion near a hospital in Gaza, nuclear power plants in Iran, and a large-scale accident on the streets of Amsterdam.
Open-source intelligence experts have emphasized that Google Earth has served as a reliable benchmark for years for verifying the authenticity of images from hard-to-reach regions.
«For decades, Google Earth has served as an outdated reference layer against which journalists and investigators have compared suspicious images. Embedding a generative tool into this very reference undermines the very measure used to detect fakes,» said Sam Stockwell, a senior fellow at the Alan Turing Institute.
OSINT expert Henk van Ess also highlighted the dangers of this approach: «The fake is now being created inside the same tool that people use to verify the authenticity of images.»
Brady Africk, an analyst at the American Enterprise Institute, noted that these fakes are tied to real satellite images, which eliminates the characteristic features of AI generation — terrain or scale mismatch. He added that watermarks are not enough, since most users do not check the details during a quick scan.
Tonight I typed just one sentence into Google Earth and put refugees near the Mexican border. Then I planted a nuclear plant in Iran. What on earth is Google doing? Check my latest post here: https://t.co/N0h6CVECqS pic.twitter.com/WSjdcEXWJx
— 𝚊𝚊𝚗𝚎 (@henkvaness) July 30, 2026
In response to criticism, Google said they were deactivating the feature to develop stricter security measures.
«We know that users rely on Google Earth as a reliable source of information about the world. We have seen geospatial experts use this feature for a variety of useful purposes, but we have also noticed that some users are sharing screenshots of generated images that appear to violate our policies. Therefore, we are temporarily disabling this feature in Google Earth while we work to implement stricter restrictions,» the developers said.
Company representatives clarified that the generated footage did not enter the main interface for other users and was labeled «AI Generated» as well as a hidden SynthID watermark.
