Google Earth pulls back on AI feature

Google Earth has removed its new feature, Google Earth Image Generation, just one day after launching it, following criticism from journalists and researchers.

Launched last week, the feature enabled users to generate AI images from simple text prompts and seamlessly integrate them with real-world satellite imagery. Google said the tool was designed to encourage creativity, but it quickly raised concerns about the spread of deepfakes.

Open-source investigators warned that the tool could be used for disinformation. Researchers demonstrated AI-generated scenes showing false disasters, including fires at Iran’s Kharg Island and flooding around the US Capitol.

Experts said satellite imagery has long been one of the most trusted tools for verifying conflicts, natural disasters, and breaking news because it is difficult to manipulate. They fear AI-generated satellite images could undermine that trust and fuel misinformation.

In a statement on X, Google said it had seen useful applications from geospatial professionals. However, it also found that some generated images appeared to violate its policies. The company said it is rolling back the feature while it implements “stronger guardrails.”