# Google Earth AI Images Let Users Fake Events on Real Satellite Maps and the Internet Is Alarmed > Source: > Published: 2026-07-31 07:33:23+00:00 # Google Earth AI Images Let Users Fake Events on Real Satellite Maps and the Internet Is Alarmed ## The fakes sit on Google's own satellite data, borrowing the trust we place in the real map Google Earth now lets anyone type a sentence and drop fake buildings, disasters, or war damage onto real satellite maps, a tool that went live worldwide this week as researchers warn it could break how Americans verify news. ## What Google Just Switched On Google turned on Nano Banana 2, its Gemini image model, inside Google Earth on the web on 30 July, letting anyone zoom to any coordinates, click 'create image,' and type what they want to see. The rollout is global, free, and carries no waitlist. For now it lives only on the desktop web version and is missing from the iOS and Android apps, whose Android version alone has topped 500 million downloads. ## Why This Isn't Just Another AI Tool The worry isn't the fakery itself. It's the grounding. Google says the model builds every picture on real satellite, aerial, and 3D data from the exact spot on screen, so an invented hospital or bomb crater sits at true coordinates in the right light and angle. Unlike a blank-box generator such as DALL-E or Midjourney, the fake inherits the credibility of the map it's drawn on. ## One Researcher's Alarm Open-source investigator Henk van Ess said he typed single sentences into the tool and planted a nuclear plant in Iran, placed refugees on a road near the Mexican border, and added a hospital and a bomb crater in Gaza, each in seconds, all on Google's own imagery. He said the job that once needed a screenshot, a second tab, and patience now needs one sentence. Writing on his Digital Digging newsletter, van Ess noted it takes a government request to pull one real building from Google Earth, but only a sentence to add a fake one. ## Satellite Proof Used To Settle Arguments For years, satellite pictures worked as hard evidence because every frame is dated. In 2014, Russia's defence ministry presented satellite images about the crash of Malaysia Airlines Flight 17 (MH17). Investigators at Bellingcat matched them against Google Earth's dated archive and found them fabricated. The reference survived that test. Now the reference itself can be rewritten. The threat is already real. As tensions ran high during last year's US-Iran conflict, the state-aligned Tehran Times posted a 'before and after' claiming to show a wrecked US base in Qatar. Agence France-Presse (AFP) and BBC Verify found the 'after' was AI-made and the 'before' was a genuine Google Earth capture of the US Navy's headquarters in Bahrain, not Qatar. Same parked cars, same rows. It still passed 950,000 views before anyone checked. Google embeds an invisible SynthID watermark in every image, and users can upload a suspect picture to the Gemini app to ask if Google made it. Yet removal tools already exist, and the Bahrain fake travelled far before fact-checkers caught up. Google pitches the feature for lighter uses, showing Pompeii as a living Roman city or sketching community gardens onto empty lots. For ordinary Americans, the stakes are simple. 'I saw it on the map' has long been a reason to trust a claim. If anyone can drop a convincing event onto real coordinates for free, that line stops proving anything. © Copyright IBTimes 2025. All rights reserved.