NewsBite analysis · AI, Cloud, Security
CNET
Google Quickly Rolls Back a Feature That Could Make Deepfake Satellite Images
Google recently introduced the Nano Banana 2 AI model that allowed users to generate images based on satellite data from Google Earth. However, due to backlash from researchers and journalists regarding the potential misuse of these realistic images in misinformation campaigns, the feature was swiftly retracted. Google acknowledged the issue and plans to implement stronger guardrails before re-launching. Despite generated images being watermarked and not visible to other users, concerns remain about the wider implications of such tools in a world where misinformation can easily spread.
MonitorConfidence: MediumShort term horizonReputational risk: High
Published 31 Jul 2026, 20:15 UTC2 min read
Visit original article ↗Lead image — CNET
Summary
TL;DR
- Google rolled back the Nano Banana 2 AI feature within a day of its launch.
- The feature allowed users to create personalized satellite images using AI.
- Concerns arose about the potential for generating misleading imagery.
- Generated images were watermarked and not publicly visible on Google Earth.
- Google promises to enhance safety measures before reintroducing its AI capabilities.
Interesting facts
- The feature was intended for personal use, but risks of misinformation were still significant.
- Google's quick response indicates sensitivity to public perception around AI technologies.