Google Earth’s AI deepfake feature was short-lived, lasting just a single day.

Google has discontinued a recently launched feature in Google Earth that allowed users to modify satellite images using AI-driven text prompts. This tool enabled the creation of AI-generated visuals of real-world scenarios; for instance, a user generated images depicting refugees near the Mexican border and a bomb crater by a hospital in Gaza. Initially, Google acknowledged that images produced with the tool included a digital watermark and emphasized that it restricted “image creation on harmful topics.”

However, just a day later, Google issued a new statement highlighting the trust users place in Google Earth for an accurate representation of the world. While the feature had been utilized by geospatial professionals for various constructive applications, the company noted that some users were sharing screenshots of generated images that breached their policies. Consequently, they decided to retract the feature while they work on implementing more robust safeguards. It’s worth mentioning that the generated images were not visible in the main Google Earth experience for other users and were marked as AI-generated.

The user who created the controversial images pointed out that he managed to bypass Hive’s AI detection tool with a video made in Google Earth, illustrating how such generated content could circulate despite the watermarks. He remarked that during his testing, “Nothing was refused, nothing was softened, and nothing suggested I try a different prompt.”