Google briefly introduced an experimental feature within Google Earth that allowed users to integrate AI-generated imagery directly onto its satellite maps. Powered by its Nano Banana 2 image generator, the tool was designed to foster creative exploration of geographical concepts. Users could employ prompts to superimpose fabricated visuals over real-world locations, aiming to offer a novel way to interact with the mapping platform.
However, the feature faced immediate and widespread criticism from various quarters, including journalists and researchers. Concerns rapidly mounted over the potential for misuse, specifically the creation and dissemination of misinformation. Critics highlighted that integrating a prompt-based AI image generator into a platform widely regarded as a reliable source of visual evidence could easily undermine its credibility and flood the internet with misleading "geospatial slop."
Responding swiftly to the backlash, Google announced the rollback of the feature just one day after its launch. The company stated that while some professionals found useful applications, instances of generated imagery violating their policies were observed. This incident underscores the critical challenge developers and researchers face in deploying powerful AI tools responsibly, particularly those capable of generating realistic content, and the ongoing need for robust guardrails to prevent the proliferation of AI-driven disinformation.