Google removed a new generative AI feature from Google Earth in less than 24 hours, reversing the launch almost immediately after users began creating misleading satellite-style images. The tool had allowed people to generate fabricated scenes, including fake natural disasters and military installations.
The fast rollback stands out because Google is usually known for extended testing periods and gradual product changes. In this case, the company appears to have acted quickly once it became clear that the feature could be used to produce convincing but false visual material inside a mapping and Earth-imagery context.
The episode highlights a broader problem facing generative AI products: when realistic-looking images are placed in environments associated with factual information, the risk of confusion rises sharply. In a platform tied to geographic data and satellite views, fake scenes can carry extra weight because users may assume the imagery reflects real events or real locations.
Google’s decision to pull the feature so quickly suggests the company saw immediate concerns around misuse and trust. The short-lived rollout is likely to add to ongoing scrutiny of how AI image tools are introduced, especially in products that people rely on for location, mapping, and visual reference information.
