Here’s The Problem With Putting An Ai Image Generator In Google Earth

Now a matter punctual is each it takes to make reality-warping images utilizing Google Earth’s satellite, aerial, and 3D imagery, for illustration these images generated by Digital Digging’s Henk van Ess that show “refugees adjacent nan Mexican border” and a explosive crater adjacent a infirmary successful Gaza.

Google responded to Digital Digging’s AI-altered images, saying, “We return misinformation earnestly – each image created pinch Nano Banana successful Google Earth includes nan SynthID integer watermark, truthful if personification is unsure astir an image, they tin inquire nan Gemini app aliases usage Lens successful Search to spot if nan image was AI-generated.” In summation to utilizing Google’s “@verifyai” tag successful Gemini to cheque suspicious images, van Ess suggests comparing them to different outer platforms, for illustration Sentinel-2 aliases Landsat, and moreover verifying orbital information for illustration erstwhile it says it was taken and by which satellite.

The characteristic launched globally connected nan web type of Google Earth earlier this week, and Google’s blog post pitches it arsenic a instrumentality for visualizing humanities sites and existent property projects. Some images are easy to spot arsenic fakes, for illustration showing nan Sphinx crossed pinch nan Statue of Liberty, but I tin spot really nan Nano Banana 2 feature could create images that are believable capable to dispersed misinformation. And while Google’s watermark whitethorn beryllium difficult to remove, it’s not intolerable to miss aliases bypass – Digital Digging was capable to fool Hive’s AI detector pinch an AI-altered video from Google Earth.