Google Earth’s AI deepfake tool only lasted one day
The Verge · 2026-07-31
Google Earth briefly launched an AI deepfake tool on July 31, 2026, which allowed users to easily alter images within the platform using AI prompts. The feature, demonstrated by a reimagining of Pompeii with "Nano Banana 2," was quickly deemed problematic and lasted only one day. This incident highlights the risks associated with providing public access to generative AI tools for image manipulation, particularly in platforms like Google Earth where geographical accuracy and historical representation are paramount. The rapid retraction suggests that the potential for misuse, such as creating misleading or fabricated imagery, was recognized soon after its release.
*The full article also explores the broader implications of AI deepfake technology and its impact on digital integrity.*