Google pulls Earth AI feature after misinformation backlash

Google has scrapped its newly launched artificial intelligence feature for Google Earth, which allowed users to generate and place fabricated images on real-world maps. The feature, introduced on Thursday, leveraged the Nano Banana 2 image generator to let people superimpose virtually any prompt-based image onto the mapping service. The company described the tool as a way to 'get creative with geography,' but critics saw it as a potential vector for misinformation.

The tool was designed to allow users to prompt the AI to create images that would then be overlaid directly on Google Earth's satellite imagery. This meant that realistic but entirely fake scenes could be placed over actual locations, raising concerns that the feature could be exploited to produce convincing visual falsehoods. Journalists and researchers often rely on Google Earth as a source of visual evidence, and one BBC journalist took to social media to sarcastically highlight the potential for abuse.

Within a day, Google reversed course. The company said that although some geospatial professionals had put the feature to productive use, it had also observed users sharing screenshots of generated imagery that appeared to violate its policies. Consequently, Google decided to temporarily disable the feature and work on 'implementing stronger guardrails' before considering a relaunch.

The episode underscores the delicate balance technology companies must strike between enabling creativity and preventing misuse. While AI-generated imagery is not new and can be produced with many tools, embedding such a feature into a widely used mapping platform elevated the risk of fabricated content being mistaken for real satellite imagery. Google's swift removal of the feature suggests an awareness of the potential reputational damage and societal harm that could result from unchecked AI-generated misinformation.