Google pulls Earth AI feature after disinformation abuse

Google discontinued its new Earth AI generative feature less than 24 hours after launch following user generation of disinformation imagery. The tool, which allowed overlaying AI-generated scenes onto real satellite data, was suspended after users created fabricated images of Iranian nuclear plants and refugees at the U.S.-Mexico border. A Google spokesperson stated that generated content was watermarked and isolated from the main public experience, but acknowledged the need to maintain trust in Earth’s reliability. While intended for urban planning and historical visualization, immediate misuse demonstrated significant safety vulnerabilities in combining generative models with authoritative mapping platforms. The suspension marks a notable setback in integrating generative AI into core utility products where factual accuracy is paramount. Google has not announced a timeline for reinstatement or specific technical mitigations required before any potential relaunch.