Google recently launched a new feature for Google Earth that allows users to generate AI-based visualizations directly on top of the platform’s satellite imagery. While the company markets this “create image” tool as a way to visualize urban development or historical changes, it has inadvertently triggered a wave of concern among open-source intelligence experts and disinformation researchers. By integrating its “Nano Banana 2” generative technology, Google has essentially enabled users to turn a once-trusted, objective tool for global verification into a canvas for high-quality, synthetic fabrication. The shift is monumental, effectively transforming a source of empirical truth into a platform that can now produce highly convincing, yet entirely manufactured, geospatial narratives.
The core of the criticism lies in the erosion of Google Earth’s reputation as the “gold standard” of geographic reference. For two decades, the platform has served as a primary source for journalists, humanitarian groups, and researchers to verify global events. Digital investigation expert Henk Van Ess pointed out the irony of this development: Google has spent twenty years building an indispensable tool for fact-checking, only to now introduce a feature that can actively manufacture falsehoods. The danger is not just in the image itself, but in the context it occupies. Because these AI-generated visuals are rendered on the interface of Google Earth, they inherit an immediate sense of institutional credibility that standalone AI images lack, making it dangerously easy for bad actors to manipulate public perception.
Experts in the field are particularly worried about the downstream effects of this technology. Brady Africk of the American Enterprise Institute notes that the ease of generating these images will likely lead to a surge in disinformation that is both rapid to spread and difficult to debunk. Beyond the immediate impact of specific fake images, there is a looming fear of “liar’s dividend”—a phenomenon where, as fakes become common, the public loses trust in all satellite imagery, including legitimate, hard-hitting evidence of war crimes or environmental disasters. By cluttering the information landscape with AI-generated noise, Google may be making the already difficult work of journalists and researchers nearly impossible to conduct effectively.
In defense of its product, Google has pointed to its safety protocols, specifically the inclusion of SynthID digital watermarks. These invisible markers are designed to signal when content is AI-generated, and Google encourages users to utilize its Gemini chatbot to verify the authenticity of an image. Furthermore, the company claims to have guardrails in place to prevent the creation of images depicting “harmful topics” or misleading content. However, these reassurances have been met with skepticism. Independent testing conducted by news agencies immediately following the launch revealed that the tool could be easily coaxed into generating images of explosions, military sites, and politically sensitive scenarios, suggesting that the “safety filters” are far less effective than the company claims.
The potential for real-world harm is not purely hypothetical; history has shown us that even poorly executed AI fakes can cause significant turmoil. From market-rattling images of explosions at the Pentagon to doctored satellite photos of military bases that gain millions of views, the public is already susceptible to visual misinformation. When these fakes are given the polished, topographical framing of Google Earth, their potential to incite panic, influence geopolitical tensions, or affect domestic election integrity is amplified. Critics like Ross Burley of the Centre for Information Resilience have been vocal in their frustration, describing the rollout as “irresponsible” and warning that the years of trust built into satellite mapping could be undone in a matter of days.
The backlash from the open-source intelligence community, including groups like GeoConfirmed, has been swift and decisive, with many calling for a total reversal of the feature. These groups have already noted a disturbing trend of users exploiting the tool to mislabel military installations or falsify the nature of infrastructure for propaganda purposes. It appears that the utility of the tool for legitimate planning is being dwarfed by its capacity for misuse. As it stands, Google is facing a critical crossroads: they must decide whether to prioritize a creative, gimmicky feature or the preservation of a platform that the world relies on to distinguish truth from fiction. If the company chooses to keep the feature, they may find that the cost of their innovation is the permanent loss of public trust in the very map that defined the digital age.