Google has rolled back a new Google Earth feature that let users generate AI images on top of its satellite imagery, following a backlash from the open-source research community, including CIR. The tool, launched last week, integrated Google's image-generation technology into Google Earth, allowing users to zoom into any location and create new pictures based on its satellite and 3D-mapping data. Google framed it as a way to visualise history and plan real estate projects. But researchers quickly warned it could be used to fabricate convincing fake satellite imagery, of explosions, military positions, or damage that never happened, on a platform that journalists and investigators rely on precisely because it is trusted. In a @FRANCE24 article, CIR Co-Founder @rwdburley called the feature "irresponsible," warning that "trust in satellite imagery has taken decades to build and could be irrevocably damaged overnight." That trust is central to open-source work - satellite imagery is one of the main tools CIR and others use to verify where and when an event took place. If audiences can no longer be confident that an image reflects reality, the cost is not just individual fakes but an erosion of the shared reference point that verification depends on. Google says it has paused the feature while it works on stronger safeguards. We welcome the decision and hope it reflects a wider recognition that tools capable of rewriting the visual record need serious guardrails before release, not after. Read on: france24.com/en/live-news/20…