Google pulled its Gemini-powered Earth image tool within 24 hours after users generated fake refugee camps and disasters, adding to AI safety scrutiny for Alphabet.
Alpha Score of 72 reflects strong overall profile with moderate momentum, moderate value, strong quality, strong sentiment.
Google pulled a new AI image-generation feature from Google Earth less than 24 hours after launch, after users exploited it to create realistic deepfakes of disasters and other sensitive scenes.
The company released the function Thursday, letting Gemini's Nano Banana 2 model modify satellite and street-level imagery based on text prompts. Within hours, AI researcher Henk van Ess generated fake refugee camps along the U.S.-Mexico border and a nonexistent nuclear plant in Iran. Other users created images of flooding and a 9/11-style attack.
“We've seen geospatial professionals using this feature for a range of useful purposes, however we've also seen people sharing screenshots of generated imagery that appear to violate our policies,” a Google spokesperson said in a statement posted on X. “So we're rolling back this feature in Google Earth while we work on implementing stronger guardrails.” The company said the images were watermarked and never appeared in the public Google Earth experience.
Google is not the first to pull an AI tool after misuse. Meta launched Muse Image on July 7 with a feature that generated images from public Instagram accounts. Meta removed it in less than a week and said it had “missed the mark.”
The incident adds to scrutiny over how Alphabet (GOOGL) and other tech giants handle AI safety. The European Union's AI Act is set to take effect incrementally, and U.S. regulators have signaled tighter rules on synthetic content. Analysts at several firms have flagged AI misuse as a potential drag on ad revenue and legal costs, though no direct financial impact has been disclosed.
The Google Earth feature was aimed at urban planners, emergency managers and educators. The company said it will re-release the tool after implementing stronger guardrails, without giving a timeline.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.