Feature Launched July 30 Allowed Single Text Prompt To Generate Photorealistic Fake Images On Real Satellite Coordinates. Google’s Own SynthID Watermark Failed To Flag Results
According to Futurism, Google launched an AI image generation tool directly inside the web browser version of Google Earth on July 30, 2026, allowing anyone to fabricate made-up satellite images of anywhere in the world using a simple text prompt then pulled it down within 24 hours after it immediately became a misinformation crisis.
Dutch OSINT researcher Henk van Ess published a Substack post documenting what single-sentence prompts produced: a nuclear facility in Iranian territory, a refugee camp at the US-Mexico border, a hospital with a bomb crater in Gaza, and a fatal crash in Amsterdam; each generated on top of authentic Google Earth satellite imagery. “The forgery does not have to look convincing on its own,” van Ess wrote. “It inherits the credibility of the map it was born on.”
Google’s own SynthID watermarking system failed to reliably flag the generated images. Bellingcat researcher Jake Godin noted that fake satellite imagery had already circulated during the US-Iran war; Iran’s Tehran Times published a fabricated “before and after” showing destruction at a US naval base in Bahrain. That image required multiple steps. The Google Earth tool collapsed that workflow into one prompt.
Google said it “knows people uniquely trust Google Earth for a reliable view of the world” and would work on stronger guardrails before redeployment.
The product launched. The red team was apparently not consulted in advance. Check out our previous coverage of tech and AI on The Trusted Times.
