South Korea as seen through Google Earth. [Google Earth screenshot]
South Korea as seen through Google Earth. [Google Earth screenshot]

Google Earth added an AI-powered image-editing feature only to pull it within a day, after critics warned it could be exploited to produce disinformation. Google Earth, a service that displays the Earth's surface using geographic data including aerial and satellite imagery, drew swift backlash over the potential for the tool to generate convincing fake images.

Google, a subsidiary of Alphabet, announced Thursday that it had integrated its AI image-generation tool "Nano Banana 2" into Google Earth. Users could zoom into specific locations using the platform's satellite, aerial and 3D map data and generate images within seconds.

Brian Horowitz, Google Earth's product manager, said the integration would enable a range of uses — visualizing and bringing history to life, drafting professional real estate plans and rendering construction projects before ground-breaking.

Open-source intelligence (OSINT) experts and disinformation researchers pushed back sharply, warning the feature was ripe for abuse. Because it could overlay plausible fake scenes onto real terrain and building imagery, they said, it would make fact-checking significantly harder. Google Earth has long served as a reference tool for journalists and researchers verifying damage in war zones, restricted sites, military facilities and disaster scenes.

According to the Financial Times, fake images generated with the new feature — including bomb craters in Gaza, Iran's nuclear facilities and a car crash on an Amsterdam street — were shared on social media platform X.

AFP said its own tests produced fake satellite images evoking internationally sensitive situations, including an explosion in Paris, Iran's nuclear facilities, bomb craters in Russia and an Islamic State training camp in Syria.

As the controversy grew, Horowitz posted an update Friday — less than 24 hours after the feature launched — announcing that the integration was being rolled back.

"We know the unique trust people have in Google Earth to show the world in a reliable way," he said. "While we saw geospatial experts using this feature for a variety of helpful purposes, we also saw people sharing screenshots of generated images that appear to violate our policies." He added that the feature would be rolled back from Google Earth "while we implement stronger safeguards."

Images created with the AI editing tool did not appear on the default Google Earth view seen by other users, and they carried a watermark identifying them as AI-generated. Experts said those measures were not enough to prevent the spread of disinformation through misuse of the tool.

Sam Stockwell, a senior researcher at the Alan Turing Institute, told the Financial Times that "Google Earth has served as a reference layer for journalists and investigators to cross-check suspicious imagery for decades." He added that embedding a generative tool within that reference material "is contaminating the very yardstick used to expose fakes."


eyre@heraldcorp.com