Google Earth abruptly reverses course on AI feature

Apparently, even mapping experts aren’t completely sure whether something’s about to go south. At least that was the case with Google Earth and the new feature it rolled out last Thursday, Google Earth Image Generation, which allowed users to generate hyperrealistic AI images and add them to actual satellite imagery. The feature lasted exactly one day.What was the point of it? Google Earth said Image Generation would be a fun way to visualize future projects, like building a house, or ponder how things might’ve looked in the past. It even showcased the feature by imagining what Pompeii looked like before disaster struck—symbolism so direct that even Nathaniel Hawthorne would scoff at it.Concern about deepfakes: Journalists immediately sounded the alarm, since Google Earth is often used to verify claims and authenticate news stories. One open-source researcher told NPR that he was able to easily create fake images that could be used for disinformation: “Refugees at the Mexican border, a nuclear plant in Iran, a crash in Amsterdam, a hospital with a bomb crater in Gaza. Nothing was refused.”Google said it pulled the feature on Friday to implement “stronger guardrails” and offered reassurances about what was already out there from the tool, stating, “It’s important to note that generated images didn’t appear in the main Google Earth experience for others to see and were watermarked as AI generated.”Zoom out: Google isn’t the only one that thought it would be fun to add an image generator to a popular platform, but couldn’t quite stick the landing. Last month, Meta pulled its AI-generation feature, Muse Image, after less than a week, following complaints that it automatically opted in all public Instagram accounts.—BC