[Tech] - Google adds AI to satellite images, raising…

On July 30, 2026, Google introduced a feature for its Google Earth platform that allowed users to generate AI-based satellite imagery simply by typing a description. While intended to spark user imagination, the tool faced immediate backlash from open-source researchers and investigators who warned that the ease of creating deepfake satellite imagery posed severe risks to information integrity and verification efforts. Google retracted the feature shortly after its launch, citing concerns over policy violations and the need for stronger guardrails.

What happened

The AI feature allowed users to zoom into any location on Google Earth and use a "create image" prompt to generate custom satellite views. Researchers demonstrated the potential for abuse by easily producing realistic but entirely fabricated events, including images of fires on Iran's Kharg Island and a flooded U.S. Capitol complex.

Experts in visual forensics and open-source intelligence, such as those from Bellingcat, expressed alarm that the tool could be used to proliferate misinformation, as satellite imagery has historically been considered a reliable "source of truth" for verifying breaking news. The ability to generate such imagery with a single click threatened to undermine the credibility of satellite data, potentially allowing actors to claim real imagery is fake to avoid accountability. While Google noted that its images included a SynthID watermark and that it had attempted to restrict generation on harmful topics, researchers argued that the tool was not adequately red-teamed. The incident highlights a growing tension between generative AI capabilities and the need for trustworthy, verifiable visual evidence in global reporting.