Google pulls AI image generator from Google Earth
Just one day after launch: Google switched off the AI image feature 'Nano Banana' in Google Earth after users created misleading scenes.
Illustration · AI-generated (AI IN LIFE)
At a glance
- Reported: 3 August 2026 (TechRadar, PetaPixel and others)
- Feature: AI image generation ('Nano Banana') in Google Earth
- Rolled back after about one day due to misuse
- Reason: misleading/faked scenes; stronger guardrails announced
Google has pulled a freshly launched AI image feature in Google Earth after about a day. The tool, based on 'Nano Banana', let users alter satellite and map views with generated content – and was promptly misused for misleading, partly faked scenes.
Reports by TechRadar, PetaPixel and others on 3 August 2026 describe how quickly users tested the limits: invented disasters, manipulated locations, sensitive facilities. Google withdrew the feature 'while we work on implementing stronger guardrails'.
The case shows the core problem of generative image AI in a mapping context: where people read maps as a depiction of reality, fakes weigh especially heavily. Trust in geodata is a valuable asset.
The lesson for all providers of synthetic media: guardrails first, launch second. Labeling, watermarks and clear limits are not an afterthought but a prerequisite – which is exactly where EU regulation is heading.
FAQ
What was rolled back?
An AI feature in Google Earth that could alter map/satellite views with generated image content.
Why?
Users created misleading and faked scenes; Google paused the feature to build stronger guardrails.
Will it return?
Google mentions work on guardrails; a return date is not confirmed.
