While Google describes the tool as a way to "bring ideas to life," tests by the source demonstrate that the AI successfully generated sensitive imagery, including a nuclear plant in Iran and a hospital in Gaza, despite company claims of safeguards against harmful topics.
This development fundamentally alters the role of Google Earth, which has served for two decades as a trusted, dated reference used by investigators and journalists to verify real-world events.
By allowing users to "weld" fabricated visuals onto genuine aerial and 3D data, the tool risks enabling high-speed misinformation and infowars by governments or bad actors.
Because these generated images inherit the lighting, angles, and credibility of the underlying map, they are difficult to distinguish from reality, potentially casting doubt on legitimate photographic evidence of atrocities or military movements.
To counter misuse, Google stated that images generated by Nano Banana 2 include SynthID, a digital watermark—an invisible marker used to identify AI-generated content.
However, this protection relies on Google’s own ecosystem, such as the Gemini app or Search, to detect the mark.
The source notes that this defense may fail when images are shared as screenshots or re-encoded videos on social media, where external AI detectors often struggle to identify the fabrications.
As a result, verifying the authenticity of satellite imagery may now require checking independent orbital data or cross-referencing multiple satellite providers like Landsat or Airbus.