Tool Withdrawal Following Expert Warnings
Google has temporarily suspended a new artificial intelligence feature within Google Earth that allowed users to generate fabricated images overlaid onto real-world satellite, aerial, or 3D imagery. The pullback followed scrutiny from BBC Verify, which reported significant concerns regarding the potential misuse of the tool and its capacity to disseminate misinformation.
Less than 48 hours after announcing the integration of Nano Banana 2, Google’s AI image generator, the tech giant stated it was pausing the feature “while we work on implementing stronger guardrails.” In a statement issued Friday, Google noted that they had observed instances of people sharing screenshots of generated imagery that appeared to violate company policies.
The restricted capability allowed users to input text prompts to create false scenarios atop genuine base images. These resulting visuals could then be downloaded and shared across various external platforms. While adding that “We know that people uniquely trust Google Earth for a reliable view of the world,” the company paused the feature after mounting expert criticism.
Concerns Over Misinformation Potential
Experts speaking with BBC Verify warned that the tool could be exploited to spread false narratives, leveraging the inherent credibility associated with Google’s platform. During testing conducted by BBC Verify, several alarming images were created using the feature, including a collapsed Eiffel Tower, a sinkhole engulfing the Great Pyramid of Giza, and depictions of Russian tanks within Ukraine’s capital.
Prior to its suspension, AI and misinformation expert Henk van Ess pointed out the high risks posed by the tool. During testing, he used it to generate fake visuals depicting non-existent sites such as a nuclear power plant in Iran, a refugee camp located on the US-Mexico border, and an artificial hospital in Gaza featuring an adjacent bomb crater.
“The forgery does not have to look convincing on its own. It inherits the credibility of the map it was born on,” van Ess stated. He noted that Google had permitted “invented” imagery to be “welded to genuine coordinates, drawn on genuine imagery.”
Technical Flaws and Erosion of Trust
Google initially claimed that all AI content generated by its tools includes invisible watermarks to identify it as manipulated or fake. The company advised users who are “unsure about an image” to check it using the Gemini chatbot app or the Lens feature within search.
However, BBC Verify’s tests revealed flaws in these safeguards. The investigation found that while Google’s checks were effective for some generated images, it was possible to circumvent these mechanisms and trick Gemini into accepting the fake Google Earth visuals as authentic. Furthermore, external AI detection tools sometimes failed entirely to identify the artificial content.
AI-detection researcher Henry Ajder warned that creating fabricated imagery in conflict zones could be particularly “destabilising” when accurate information is scarce or events are moving quickly. He added that while journalists may have the capacity to verify such images, not everyone will take the time to do so.
Ajder also highlighted instances where Google’s initial guidelines were bypassed. While a request for a “raised platform to hang traitors” next to the UK parliament was rejected, using less specific language resulted in acceptance. Similarly, while an attempt to digitally remove the name “Trump” from the White House lawn failed according to guidelines, generating a community garden in that area was permitted.
“It’s concerning because it’s another place where sources of trusted information are being eroded by the constant doubt that AI might be involved,” Ajder remarked. He emphasized that areas perceived as “unblemished by AI are valuable, particularly for journalists and people trying to do work on the frontlines of rapidly evolving crisis scenarios.”
Bill Greer, a geospatial analyst and co-founder of the non-profit satellite service Common Space, noted that satellite imagery has historically been considered an exceptionally trustworthy data source. He cautioned that because falsified imagery is already used to spread misinformation, combined with governments increasingly restricting data access, there will likely be more false images and fewer genuine ones, which will significantly undermine public faith.
