Google has introduced a controversial new feature to Google Earth that allows users to generate AI-enhanced visualizations directly over existing satellite imagery. Powered by the company’s “Nano Banana 2” technology, the tool is designed to help users model real estate developments or visualize historical changes. However, the rollout has triggered an immediate backlash from open-source intelligence (OSINT) experts and disinformation researchers, who argue that integrating generative AI into a platform long considered the “gold standard” for objective geographical truth undermines the integrity of global investigative work.
The core concern among experts is the “inheritance of credibility.” Digital investigation specialist Henk Van Ess noted that because these AI-generated images are embedded within the Google Earth interface, they acquire an automatic sense of authority, making it difficult for the average viewer to distinguish between actual satellite data and manufactured fiction. Research analyst Brady Africk of the American Enterprise Institute warned that this move could accelerate the spread of sophisticated visual misinformation, creating a ripple effect that damages public trust in satellite imagery and complicates the verification efforts of journalists and conflict analysts worldwide.
In response to the criticism, Google has emphasized its commitment to safety, noting that all AI-generated content includes “SynthID” digital watermarks—invisible markers that can be detected by the company’s Gemini chatbot. The company stated that it has safeguards in place to prevent the generation of images related to harmful topics and strictly prohibits the use of its AI tools for “misinformation, misrepresentation, or misleading activities.” Google maintains that it takes the threat of disinformation seriously and is actively monitoring the use of the new feature to ensure compliance with its policies.
Despite these assurances, testing performed by international news agencies has demonstrated the ease with which the tool can be exploited to create high-stakes, realistic fakes. Within hours of the launch, users were able to generate fabricated scenes of major global events, including an explosion in Paris, a nuclear facility in Iran, bomb craters in Russia, and staged military training camps in Syria. Furthermore, the tool successfully created misleading imagery related to sensitive domestic issues, such as migrant caravans at the U.S.-Mexico border and suspicious activity in U.S. election battleground states, illustrating the potential for the technology to influence political discourse.
The real-world consequences of such misinformation are already well-documented, with previous AI-generated hoaxes, such as an image depicting an explosion at the Pentagon, having caused brief but significant turbulence in financial markets. Ross Burley, executive director of the Centre for Information Resilience, labeled the release “irresponsible,” arguing that the trust built over decades of satellite-based transparency is now at risk of being irreparably damaged. Burley and his organization are among those formally calling for Google to roll back the update, citing the potential for the tool to be weaponized in the ongoing global information war.
The backlash has extended to the OSINT community, with the prominent volunteer-driven group GeoConfirmed reporting that individuals are already using the tool to manipulate the information environment, including relabeling civilian buildings as military targets. The group concluded that the tool’s primary utility appears to be the fabrication of evidence rather than legitimate planning or historical visualization. As the debate intensifies, the incident highlights a growing tension between the tech industry’s push for generative AI integration and the urgent need to maintain verifiable, reliable sources of truth in an increasingly digitized global landscape.
