Google has removed a newly launched Google Earth feature after roughly one day of public availability, following warnings from disinformation researchers that it allowed users to overlay convincing fabricated scenes onto the platform's real satellite imagery.

The feature drew heavy criticism from multiple open-source intelligence professionals who called for its immediate removal. Google initially attempted to implement safeguards to keep the tool operational but ultimately conceded and pulled it.

How the Feature Worked

The feature, called "create image," used Google's Nano Banana 2 model to transform typed text prompts into invented visual scenes layered over genuine satellite, aerial, and 3D map data. Product manager Bryan Horowitz said users could navigate to any location on the web version of Google Earth, tap a button, and type whatever they wanted to appear there. Google marketed the tool as a way to "visualize history, create real estate plans and more."

However, open-source investigator Henk van Ess wrote on Substack that Google had spent two decades building a trustworthy geographic reference, only to add a "button that makes things up." His central concern was that AI-generated images would borrow the credibility and authority of the Google Earth platform itself — a platform widely used by journalists, humanitarian organizations, and government agencies for conflict monitoring, disaster assessment, and geographic verification.

Widespread Testing Reveals Flaws

Multiple news organizations and investigators quickly demonstrated the tool's potential for abuse. BBC Verify used the tool to generate images of a collapsed Eiffel Tower, a sinkhole swallowing the Great Pyramid of Giza, and Russian tanks in Kyiv. Van Ess separately created images depicting refugees at the Mexican border and a bomb crater beside a hospital in Gaza — neither of which triggered any safeguards.

NPR produced images showing Iran's Kharg Island in flames and a flooded U.S. Capitol. AFP fabricated images of an explosion in Paris and a nuclear site in Iran. 404 Media staff added skyscrapers to farmland, planted a "homeless encampment" in Los Angeles, and staged a protest outside Google's own offices.

Why Satellite Imagery Credibility Matters

Open-source analysts rely on Google Earth imagery precisely because authentic photographs taken from hundreds of miles overhead are costly and technically difficult to forge. Jake Godin, a senior researcher at the investigative group Bellingcat, noted that fabricated satellite images have appeared before — including fake imagery purporting to show damage at a U.S. base in Bahrain during the U.S.-Israel conflict with Iran. However, he explained that introducing a one-click generator makes such fakes far too easy to mass-produce.

Bill Greer, a geospatial analyst and co-founder of the nonprofit Common Space, described the erosion of trust in satellite imagery as "potentially particularly damaging." The concern extends beyond professional investigators: satellite imagery shared on social media is often treated by the public as inherently trustworthy, and a flood of convincing fakes could make it harder to distinguish genuine evidence of real-world events from fabrications.

Google's Safeguards Proved Insufficient

Google stated that every image created with the tool carried a SynthID digital watermark, verifiable through the Gemini app or the Lens feature in Search, and that it did not permit image creation related to harmful topics. However, BBC Verify subsequently found it could trick Gemini into confirming a fake Google Earth image as genuine. SynthID is part of a broader push across the AI industry toward content provenance standards, but the finding underscored the difficulty of making watermark-based detection reliable when a platform's own verification tools can be misled.

The platform's prompt filtering also proved easily circumvented. While it rejected a request to build a gallows to "hang traitors" outside the UK Parliament, a vaguer version of the same prompt passed through without issue.

Google has since removed the feature, acknowledging that while geospatial professionals found the tool useful, users were sharing screenshots of generated imagery that violated the company's policies. Google stated it will work on implementing stronger guardrails, though whether the tool will return remains an open question.