10
Loading video...
By Alice Gibbs
Senior Reporter
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
Google has pulled its Nano Banana image-generation tool from Google Earth less than a day after launch, following widespread concerns that users were using the feature to create realistic but fabricated images of disasters, military activity and other potentially misleading events.
"It's Chernobyl-disaster-level harmful in terms of misinformation," Nathaniel Raymond, executive director of Yale School of Public Health's Humanitarian Research Lab, told Newsweek. "Google Earth is not a social media platform. It is a forensic record."
Following the launch, social media users quickly demonstrated how the tool could be used to insert false scenes into trusted satellite imagery, including apparent disaster zones, conflict scenarios and other fabricated events. Examples shared online showed how easily convincing screenshots could be produced and circulated outside the Google Earth platform.
Read More on News
...
Google Reverses Course
Google confirmed on July 31 that it was rolling back the feature while it worked on stronger safeguards.
“We know that people uniquely trust Google Earth for a reliable view of the world. We’ve seen geospatial professionals using this feature for a range of useful purposes, however we’ve also seen people sharing screenshots of generated imagery that appear to violate our policies," the company said in a statement. "So we’re rolling back this feature in Google Earth while we work on implementing stronger guardrails. It's important to note that generated images didn’t appear in the main Google Earth experience for others to see and were watermarked as AI generated.”
The rapid reversal came after mounting criticism from researchers, open-source investigators and misinformation experts, who warned that the technology could undermine trust in one of the world's most widely used mapping platforms.
Google declined to comment further when approached by Newsweek.
...
Why Experts Were Alarmed
Among the most vocal critics of Nano Banana was Raymond, whose lab uses satellite imagery to investigate alleged war crimes, mass atrocities and humanitarian crises around the world.
He argued that Google Earth occupies a unique position because it functions not simply as a mapping product but as a widely used archive of high-resolution satellite imagery that researchers, journalists, investigators and courts rely on as a record of real-world events.
"Google Earth is not a social media platform in the traditional sense," Raymond said. "It is a forensic record."
For Raymond, the biggest concern was not that AI-generated images could be created, but that they could be presented within the familiar Google Earth environment, potentially making fabricated scenes appear more credible.
He said his team and other geospatial investigators already spend significant time combating attempts to undermine or misrepresent satellite imagery used in investigations. Introducing AI-generated content into the platform, he warned, could dramatically increase that challenge.
"It would open a fire hose of misinformation," Raymond said, explaining that investigators could be forced to spend countless hours verifying whether widely shared screenshots represented authentic imagery or AI-generated fabrications.
Raymond said the issue has implications far beyond internet hoaxes. His lab has used Google Earth imagery as a baseline for investigations into conflicts and alleged human rights abuses, including research examining sites linked to the relocation of Ukrainian children into Russia.
"This is misinformation that kills people," he said. "The issue with imagery is that it creates the perception of objective, scientific, empirical truth."
Raymond also described the rollout as a "breach of trust" for many in the geospatial community, saying researchers and mapping experts have worked alongside Google Earth for years to help maintain the integrity of a platform that has become an important public archive of satellite imagery.
Social Media Users Quickly Tested the Limits
Almost immediately after the feature launched, users began posting examples on Threads and other platforms showing fabricated scenes placed onto real-world locations.
One widely shared post demonstrated how the model could be used to create convincing disaster imagery, with images appearing of the 9/11 terror attacks, Paris’ Eiffel Tower collapsed on the ground, and even a large hole where the Great Pyramid of Giza should be.
Other examples of the feature's use included pictures of fictitious war cemeteries and changes to military sites around the world.
...
What Nano Banana Was Supposed to Do
Google introduced the integration as a creative and educational tool. The company showcased examples such as reconstructing ancient Pompeii, visualizing future building projects and creating informational graphics tied to landmarks including the Statue of Liberty.
Users could navigate to a location in Google Earth and then describe how they wanted the scene to appear, with the AI generating a new image based on the underlying geography.
Google Developing ‘Stronger Guardrails’
Google says the tool has been removed from Google Earth while the company develops stronger guardrails.
The episode is likely to become an early test case in the growing debate over how AI-generated imagery should be integrated into products that users traditionally regard as authoritative sources of visual information.
As generative AI continues to improve, experts warn that the challenge is no longer whether realistic fake images can be created, but whether platforms can prevent them from being mistaken for reality.
Contact Newsweek editors on this story: Rebecca Flood and Emma Lee-Sang
View 0 Comments
Submit Correction
View Editorial & AI Guidelines