Google removed the Nano Banana integration from Google Earth just two days after introducing it. The feature allowed users to generate custom satellite scenes using AI, enabling edits such as placing new buildings on empty lots or altering historical ruins. Within hours, individuals created convincing fakes depicting a refugee column at the Mexican border and a bombed-out hospital in Gaza. Although Google stated that these images were labelled as AI-generated and remained invisible to other users within the platform, screenshots shared publicly violated company policies. The tool will not return until stronger safeguards are implemented, as Google acknowledged the unique trust people place in its mapping services.
The incident highlights how quickly generative capabilities can be weaponised against a trusted platform. While anyone can manipulate screenshots with standard editing software, Google made the process exceptionally simple by embedding the functionality directly into the application. This lowered the barrier for creating misinformation that could damage credibility or incite unrest. The rapid removal demonstrates an admission that the initial safety measures were insufficient for the specific risks involved.
* The feature allowed real-time generation of specific geopolitical scenarios.
* Public sharing of screenshots bypassed internal visibility restrictions.
* Google confirmed the tool is offline pending improved safety controls.
