Tech giant Google has temporarily pulled a newly launched Artificial Intelligence (AI) feature from Google Earth just one day after its release. The sudden rollback was prompted by users leveraging the tool to generate and share images that violated the company’s internal content policies.
Powered by Google’s Nano Banana 2 AI model, the feature allowed users to enter text prompts to generate highly realistic imagery of any location worldwide, utilizing satellite, aerial, and 3D data from Google Earth.
Key Highlights of the Rollback:
- Official Company Statement: Brian Horowitz, Product Manager for Google Earth, stated that users rely on Google Earth for a trusted view of the world. While geographers and researchers used the feature productively, policy-violating imagery led to an immediate suspension until stronger safety guardrails are implemented.
- Safety Protocols in Place: Google confirmed that all AI-generated images featured visible watermarks to differentiate them from actual imagery and were not exposed to other users within the standard Google Earth experience.
- Path Forward: Google is actively refining safety protocols and guardrails to prevent misuse before re-introducing the generative feature in a future update.
📊 Fact-Sheet Summary
- Company: Google (Alphabet Inc.).
- Platform Affected: Google Earth.
- Feature: Generative AI imagery based on text prompts (Powered by Nano Banana 2 AI model).
- Reason for Rollback: User policy violations and low-quality/inappropriate image generation.
- Duration Active: Withdrawn after just 24 hours of release.
- Key Official: Brian Horowitz (Product Manager, Google Earth).
- Current Status: Feature paused pending enhanced safety guardrails.
