Google yanks Google Earth AI image tool after 1 day over fake satellite fears

Google introduced an AI image-generation feature inside Google Earth on July 31, 2026, leveraging its Nano Banana 2 model. The tool let users overlay or manipulate satellite imagery with AI-generated visualizations, ostensibly for creative or planning use cases. But within 24 hours, the company retracted the feature entirely, citing misinformation risks after a wave of criticism from researchers and AI safety experts.

Critics argued that Nano Banana 2's ability to synthesize realistic terrain, buildings, and weather patterns on demand could produce convincing fake satellite images that are nearly impossible to distinguish from real ones. With conflicts and disaster response already relying heavily on satellite intelligence, the potential for fabricated evidence — from false troop movements to doctored flood damage — posed a clear threat. Google's decision to pull the tool on August 1 marks one of the fastest feature reversals in its history, underscoring how even capable generative models face heightened scrutiny when they intersect with geospatial truth.