Google removed an AI image-generation feature for satellite images on Google Earth after concerns it could create deepfake maps [1].
The swift reversal highlights the growing tension between rapid AI deployment and the potential for large-scale disinformation. Because satellite imagery is often viewed as objective proof of ground reality, the ability to synthetically alter these maps could undermine humanitarian efforts or geopolitical stability.
The tool was launched on July 30, 2026 [3]. However, it was paused or removed by July 31, 2026 [2], meaning the feature existed for less than 24 hours [4].
Researchers, journalists, and human-rights advocates warned that the tool could enable the creation of deepfake satellite maps. These experts said the technology could make it easy to spread misinformation and abuse the platform's perceived authority [1, 5].
The feature was integrated into Google Earth’s online satellite-imagery platform [1, 2]. While the tool aimed to provide generative capabilities, the risk of creating fraudulent evidence of disasters or military movements prompted the immediate shutdown [5].
Google has not provided a timeline for whether the feature will return with stricter safeguards. The incident underscores a recurring pattern in the industry where generative tools are released to the public before the potential for misuse is fully mitigated [1].
“Google removed an AI image-generation feature for satellite images on Google Earth after concerns it could create deepfake maps.”
This event demonstrates the fragility of 'visual truth' in the age of generative AI. By attempting to apply AI image generation to geospatial data, Google touched a sensitive nerve regarding the verification of physical events. The rapid removal suggests that the risk of enabling state-sponsored or malicious disinformation outweighs the commercial utility of the feature, signaling a more cautious approach to AI integration in mapping services.
