Google has shut down a Google Earth feature launched Thursday that allowed users to edit satellite images with text prompts using AI. The tool essentially let users create AI-generated deepfakes of real-world locations. Digital Digging's Henk van Ess demonstrated the vulnerability by generating images adding refugees near the Mexican border and a bomb crater by a hospital in Gaza.
Google initially responded that images included a digital watermark and the company prevented "image creation on harmful topics." However, one day later, Google reversed course, stating it had "seen people sharing screenshots of generated imagery that appear to violate our policies" and was rolling back the feature while implementing "stronger guardrails."
Van Ess also revealed he was able to fool Hive's AI detection tool with a video generated in Google Earth, demonstrating how watermarked content could still spread undetected. He noted that during testing, "Nothing was refused, nothing was softened, and nothing suggested I try a different prompt."