Google removed the Nano Banana 2 AI image-generation feature from Google Earth on Friday, July 31, 2026 [1, 4].
The decision follows immediate concerns that the tool could be used to fabricate realistic satellite imagery, potentially fueling global disinformation campaigns. Because satellite data is often viewed as an objective source of truth for verifying ground events, the ability to generate synthetic versions of this data poses a significant security risk.
Google integrated the Nano Banana 2 tool into its online satellite-imagery service to allow for AI-generated visual modifications [1, 2]. However, the feature quickly drew backlash from experts and users who said the technology could be abused to create deep-fake satellite images [3, 4].
Critics argued that the tool lacked sufficient safeguards to prevent the creation of deceptive visuals. One source said, "The opportunities for abuse and disinfo are literally boundless" [4].
Google responded to the outcry by rolling back the feature shortly after its release [1, 3]. The company's action reflects a growing tension between the push for generative AI integration and the need to maintain the integrity of geospatial data.
While Google has not detailed a specific timeline for a revised version of the tool, the rollback serves as a cautionary example of the risks associated with applying generative AI to mapping and surveillance tools [2, 5].
“The opportunities for abuse and disinfo are literally boundless.”
This incident highlights the vulnerability of geospatial intelligence in the era of generative AI. As satellite imagery is frequently used by governments and journalists to verify military movements or human rights abuses, the existence of a tool that can create 'deep-fake' landscapes threatens the reliability of remote sensing as a verification tool.
