Global Markets
Home›Global Markets›North America›Google Earth disables an AI deepfake imagery feature a…
Google Earth disables an AI deepfake imagery feature after backlash
The tool was briefly available for about 24 hours, letting users type instructions to generate realistic satellite-style images that researchers said could be used for sensitive scenarios.
For about 24 hours beginning on Thursday, Google Earth offered a new artificial intelligence feature that let users generate deepfake aerial images by entering a short prompt, according to The New York Times Business.
The Times reported that a six-word instruction, “type whatever you want to see,” made it possible for billions of users worldwide to create realistic imagery, and that it drew scrutiny after open-source intelligence researchers, journalists, and others demonstrated examples including nuclear power plants, refugee camps, burned forests, flooded cities, and a fatal car crash.
Google disabled the feature on Friday, one day after announcing it, after the company said it had seen people share screenshots of generated imagery that appeared to violate its policies.
The Times noted that Google said it had also seen geospatial professionals using the feature for useful purposes, while the broader risk of disinformation pushed the company to turn it off.