Google has officially suspended a controversial new feature on Google Earth that allowed users to generate AI-driven visualizations atop real-world satellite imagery. The tool, which utilized Google’s “Nano Banana 2” technology, was intended to help users visualize urban planning and historical changes. However, shortly after its Thursday rollout, the company faced an intense firestorm of criticism from researchers, intelligence analysts, and open-source investigators who argued that the feature served as a “factory” for high-stakes digital disinformation.
The primary concern among experts was the threat posed to the integrity of geospatial data. Google Earth has long served as a gold-standard reference point for journalists, human rights organizations, and international monitors attempting to verify global events. By introducing an AI tool capable of synthesizing realistic fake imagery, critics argued that Google was effectively undermining the very foundation of truth it had spent two decades building. Digital investigations expert Henk Van Ess lamented that the company had essentially added a “button that makes things up” to a platform globally relied upon for objective reality.
The potential for real-world harm was demonstrated during independent tests of the tool, which proved capable of generating convincing depictions of catastrophic events, such as explosions in major cities, fabricated military activity in conflict zones, and falsified scenes at sensitive political borders. Analysts from the American Enterprise Institute and organizations like GeoConfirmed warned that bad actors were already exploiting the feature to create fake military positions or manipulate narratives surrounding election integrity. These experts cautioned that such AI-generated content could easily be weaponized to incite social unrest or rattle financial markets, echoing the panic caused by previous AI-hoax incidents, such as the widely circulated fake image of a Pentagon explosion in 2023.
In its defense, Google initially attempted to mitigate concerns by highlighting the inclusion of “SynthID” digital watermarks, which are invisible markers designed to identify AI-generated media. The company stated that it took misinformation seriously and directed users to its existing policies against misleading activities. However, the sheer ease with which the tool could be used to manufacture high-impact fabrications left many observers unconvinced that technical safeguards were sufficient to prevent widespread abuse. Organizations like the Centre for Information Resilience described the release as “irresponsible,” arguing that public trust in satellite intelligence could be decimated in a matter of hours.
The rapid backlash eventually forced Google’s hand. By Friday, the “create image” button had been completely removed from the Google Earth interface. In a statement to AFP, a company spokesperson acknowledged that while professionals had found utility in the feature, the company had observed users violating policies by sharing screenshots of manipulated imagery. Google confirmed it is currently pulling the feature back to work on “stronger guardrails,” prioritizing the preservation of the service’s status as a reliable window into the physical world.
The incident serves as a significant cautionary tale in the ongoing debate over the integration of generative AI into established public platforms. It highlights the inherent tension between technological innovation and the necessity of maintaining a “source of truth” in an increasingly polarized digital landscape. As researchers continue to call for more robust oversight, the rollback underscores that even the most well-intentioned tools can be disastrous if they inadvertently facilitate the mass production of synthetic reality, potentially causing irreversible damage to the public’s ability to discern fact from fiction.
