"There's No Way This Could Be Abused": Google Pulled Its AI Earth Feature in Under 24 Hours — and the Sarcasm That Killed It

On Thursday, July 30, 2026, Google launched what its product manager Bryan Horowitz described as a way to "spark user imagination" — an AI image generation tool embedded directly into Google Earth, allowing anyone in the world to zoom to any location, click a "Create Image" button, type a prompt, and have the Nano Banana 2 model produce a photorealistic image built from that location's actual satellite and aerial data. On Friday, July 31 — less than twenty-four hours later — Google pulled it entirely.

A BBC journalist posted sarcastically on Thursday: "There's no way that this new AI image generation feature on Google Earth, one of the most reliable sources of visual evidence for journalists and researchers, could possibly be abused to spread misinformation online." By Friday, it had been proven right. 

What the Tool Actually Did

Product manager Bryan Horowitz described the "Create Image" button in his launch post: zoom to any location, type a prompt, and Nano Banana 2 returns a photorealistic image built from Google's satellite, aerial and 3D imagery of that exact spot. The post pitched five uses: a classroom reconstruction of Pompeii, an infographic about a landmark, a real-estate concept for an empty Tokyo lot, a cabin rendered into a real lakefront, and a science fiction scenario. Availability was global on day one, with no waitlist, no application, and no enterprise-only tier. 

The vision, in isolation, was not unreasonable. The ability to visualize what a location looked like before a historical event, or to help students understand geographic transformation, or to assist planners in visualizing development concepts, represents a genuinely useful category of tool. Google presented it as such.

The problem was everything else a tool with those capabilities could also do.

What Journalists and Researchers Found

Investigative journalist Henk van Ess and newsrooms including NPR and Futurism proved within hours that the tool could be used to generate potentially misleading content. Tests produced false depictions of explosions, military sites, floods, and other events. Forbes reported the tool was used to generate a fake image of a nuclear plant in Iran. Fabricated satellite imagery circulated well before this launch — including an AI-altered image of a US Navy base in Bahrain that ran in Iranian media earlier in 2026 — and the tool made that process dramatically faster and more accessible.

Jake Godin, a senior researcher at Bellingcat, told NPR the change was one of speed rather than of kind: "It's just streamlining the process, which I think is going to make it proliferate more." That sentence captures precisely why the open-source intelligence and geospatial verification communities reacted so quickly and so forcefully: fabricated satellite imagery already existed as a disinformation tool. What Google had built was a one-click version of it, available to anyone, anchored to the interface that the world's journalists, human rights researchers, and fact-checkers treat as the authoritative reference layer for visual evidence. 

Google Earth's archive is dated, its coverage is broad, and newsrooms, human-rights researchers and open-source investigators treat it as the reference layer when a photograph appears claiming to show a strike, a camp or a flood. That is why the rollback statement opens on trust rather than on capability. The product being defended is the reference, not the generator.

Google's Statement and the Rollback

Google said on July 31: "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. So we're rolling back this feature in Google Earth while we work on implementing stronger guardrails."

The company added a technical note: generated images didn't appear in the main Google Earth experience for others to see and were watermarked as AI generated. That is true — but it was immediately and correctly identified as insufficient, for two reasons. First, watermarks on images shared as screenshots are trivially removed or simply cropped out. Second, the incident demonstrates that visible watermarks and platform-level labels may not be enough when AI-generated content can easily be copied and distributed elsewhere. 

Google did not provide details about the restrictions it plans to introduce. Possible safeguards could include tighter limitations on prompts involving conflicts, disasters, sensitive facilities and public safety incidents. The company has not announced a timeline for a possible relaunch. 

A Pattern Google Has Been Here Before

In February 2024, Google paused Gemini's image-generation tool shortly after its launch following user reports of historical inaccuracies. The AI Overviews feature in Google Search was also rolled back after producing dangerous incorrect advice. Each incident has followed roughly the same arc: rapid global launch, rapid discovery of a serious misuse case, rapid rollback, and a promise of stronger guardrails before relaunching. 

The swift rollback reflects the trial-and-error costs of deploying AI products. Halting the feature within 24 hours of launch shows the direct impact of public pressure on product decisions, and also exposes the lack of fact-checking mechanisms when generative AI is combined with real-world data. 

The specific combination at work here — a model capable of generating photorealistic images anchored to real geographic coordinates, distributed globally without restriction — was one that, in retrospect, the company should have been able to anticipate would be stress-tested within hours. The geospatial intelligence and OSINT communities have spent years developing expertise in satellite imagery analysis precisely because fabricated or manipulated geographic imagery is one of the most dangerous and least verifiable categories of disinformation. Building a tool that makes generating that imagery dramatically easier and embedding it into the world's most trusted map — then launching it globally before any significant safety testing had been done publicly — was not a failure of capability. It was a failure of anticipation.

For developers and enterprise users, the event sends a signal: embedding AI image generation into map-based platforms requires prioritizing content moderation and ethical boundaries. Otherwise, similar tools can easily be used in scenarios involving natural disasters, urban planning, or public opinion manipulation, eroding overall platform trust. 

The sarcasm from the BBC journalist's post did what no internal review apparently did: it identified the obvious problem in the time it takes to read a single sentence. Google Earth is back to showing the world as it actually is. For now, that is enough.

Sources

1. Yahoo Tech / AFP — Google forced to roll back new AI tool just hours after launch
2. Forbes — Google Pulls Earth's AI Image Tool a Day After Launch, Amid Nuclear Plant Fake
3. TechCrunch — Google nixes its Earth AI feature one day after launch, amid criticism it would spread misinformation
4. Unite.ai — Google Pulls Earth's AI Image Tool a Day After Launch
5. TechXplore — Google rolls back new satellite image AI tool after backlash
6. entARABI — Google Pulls AI Image Generator from Google Earth Hours After Launch
7. Winzheng — Google Suspends Google Earth AI Image Generation Feature After Deepfake Satellite Image Risks Trigger Emergency Rollback