Google is facing a significant backlash following the integration of its new generative AI model, “Nano Banana 2,” into the Google Earth platform. While the tech giant positioned the tool as a helpful utility for urban planners, researchers, and real estate developers to visualize landscape modifications and historical reconstructions, the rollout has triggered alarm among digital safety experts. Critics argue that the ability to layer synthetic, photorealistic imagery over real-world satellite data introduces a dangerous new vector for geopolitical misinformation.
The core of the controversy lies in the tool’s capability to generate highly convincing, manipulated visuals tied to authentic geographical coordinates. Previously, fabricating satellite imagery required a sophisticated level of technical expertise in digital editing; however, Google’s latest feature democratizes this process by allowing users to create altered landscapes through simple text prompts. Researchers warn that because these fabrications are “welded” to genuine map data, they become indistinguishable from reality to the average user.
Disinformation experts and open-source intelligence (OSINT) analysts have voiced deep concerns regarding the lack of sufficient protective safeguards. Online research expert Henk van Ess, who conducted hands-on testing of the feature, emphasized that the seamless integration of “invented” imagery onto real terrain data creates a high risk of deception. His findings suggest that even when using external AI detection tools, many of the images generated by the software were difficult to identify as synthetic, posing a major challenge for verification efforts.
The potential implications of this technology are particularly severe in the context of active global conflicts. AI-detection researcher Henry Ajder, speaking to BBC Verify, highlighted that the release of such tools during volatile, fast-moving geopolitical events could be incredibly destabilizing. In scenarios where reliable, ground-truth information is already scarce, the rapid spread of convincing fake satellite imagery could easily be weaponized to distort public perception or incite unnecessary panic.
In response to the mounting criticism, Google has defended the safety of the platform, noting that all AI-generated content is embedded with invisible watermarks designed to identify the images as synthetic. The company maintains that these markers provide a layer of accountability for users who share content across social media and other digital platforms. Furthermore, Google has encouraged users who remain skeptical of an image’s authenticity to utilize built-in verification methods, such as the Gemini chatbot app or Google Lens, to cross-reference visual data.
Despite these assurances, the debate underscores the broader tension between technological innovation and the ethical responsibilities of major tech companies. As AI-powered image generation becomes more advanced and accessible, the burden of truth verification is increasingly shifting onto both the platforms that host these tools and the public that consumes the content. Until more robust protections are established, the integration of generative AI into mapping services remains a controversial development that many fear could fundamentally compromise the integrity of digital geographical information.
