Google Unveils AI-Powered Satellite Imagery Tool, Raising Concerns Among Researchers
In a recent announcement, Google introduced a feature that empowers users of Google Earth to generate artificial intelligence (AI) visualizations atop the platform’s satellite imagery.
This development has triggered alarm among researchers and experts in open-source intelligence, who are apprehensive about its potential for fostering disinformation.
The newly launched “create image” tool, which debuted on Thursday, utilizes Google’s Nano Banana 2 image-generation technology.
It enables users to zoom into specific locations and fabricate images in mere seconds, drawing from Google Earth’s satellite, aerial, and three-dimensional mapping data.
Marketed as a means to “visualize history, create real estate plans, and more,” this feature has led disinformation scholars to caution that nefarious actors could exploit Google Earth—a crucial resource for validation and verification—to produce realistic geospatial forgeries.
Digital investigations authority Henk Van Ess articulated his concerns on Substack, stating, “Google spent 20 years building the reference the world checks against.
Today it added a button that makes things up.” He further noted that “the forgery does not have to look convincing on its own; it inherits the credibility of the map from which it originates.”
Brady Africk, a research analyst at the American Enterprise Institute who specializes in satellite imagery analysis, emphasized to AFP that the Google Earth update “facilitates the generation of credible fake satellite imagery, which can disseminate rapidly online and mislead the public.”
He warned that these increasingly prevalent AI-generated fakes may not only undermine public trust in satellite imagery but also complicate the work of journalists and researchers alike.
In a response to AFP via X, Google asserted that images produced using this new feature would include SynthID digital watermarks—subtle markers embedded in media generated through the company’s AI tools. Users may also inquire with Google’s Gemini chatbot regarding the presence of a SynthID.
Google emphasized its commitment to counteracting misinformation, stating that it takes these concerns seriously.
Moreover, the company declared its measures to prevent “image creation on harmful topics and its AI policy that discourages misinformation, misrepresentation, or misleading activities.”
However, tests conducted by AFP on the feature revealed the capacity for generating satellite images with significant global implications—including a fictional explosion in Paris, a nuclear facility in Iran, a bomb crater in Russia, and an alleged ISIS training camp in Syria.
This tool also successfully generated prompts that depicted fabricated scenarios such as widespread flooding in Bangladesh, a migrant caravan at the US-Mexico border, and a dubious warehouse in Georgia—one of several states where former President Donald Trump has baselessly claimed election fraud.
Previously, similar fictions have led to substantial real-world repercussions. For instance, in 2023, an AI-generated image portraying an explosion at the Pentagon briefly unsettled financial markets.
Similarly, an altered satellite image suggesting the destruction of a US military base during the Iran conflict amassed millions of views across various platforms.
Ross Burley, executive director of the London-based Centre for Information Resilience, expressed grave concerns to AFP, stating, “I cannot stress how irresponsible this is. Trust in satellite imagery has taken decades to build and could be irrevocably damaged overnight.”
Burley highlighted that the organization is advocating for Google to rescind this decision. Furthermore, GeoConfirmed, a volunteer-driven open-source intelligence initiative dedicated to verifying visuals from global conflicts, has also urged Google to reconsider its actions.
The organization noted on X, “We are already observing individuals exploiting the new Google Earth web tool to construct military positions, mislabel buildings as schools, and experiment with manipulating the information landscape.”
They concluded, “The primary, and perhaps only, practical use of this tool appears to be the creation and dissemination of misinformation and disinformation.”
Source link: English.aawsat.com.
