According to TechCrunch, Google launched an AI image-editing feature in Google Earth on Thursday, letting users generate fabricated satellite imagery with Nano Banana 2. The Verge reports Google rolled it back about a day later after researcher Henk van Ess showed it could add refugees and a bomb crater near a Gaza hospital, and after a BBC journalist warned of misinformation risks.
When Did Google Launch and Then Kill the Earth AI Editing Feature?
Google rolled out a new feature on Thursday that let users deploy Nano Banana 2, its AI image generator, to create fabricated images inside Google Earth, its satellite imagery mapping app, according to TechCrunch (E6). The Verge similarly reported that "Google has shut down [a] Google Earth feature it launched Thursday that allowed users to edit satellite images with text prompts using AI" (E1).
The turnaround was fast. TechCrunch noted that "only a day after the feature's release, Google has scrapped it" (E9), and The Verge's account of the shutdown timeline matches that same one-day window from launch to rollback (E1).
What Safeguards Did Google Initially Put in Place — and Why Weren't They Enough?
When researcher Henk van Ess first flagged problems with the feature, Google's initial response noted that images generated with Nano Banana 2 in Google Earth included a digital watermark and that the company prevented "image creation on harmful topics," according to The Verge (E3).
That defense didn't hold up under testing. Van Ess said he was able to fool Hive's AI detection tool with a video generated in Google Earth, demonstrating that generated media could spread despite the watermarks. In his own testing, he said, "Nothing was refused, nothing was softened, and nothing suggested I try a different prompt" (E5) — directly contradicting Google's claim that harmful-topic generation was being blocked.
What Fabricated Content Did People Actually Generate With the Tool?
According to The Verge, van Ess "intentionally generated images adding things like refugees near the Mexican border and a bomb crater by a hospital in Gaza" (E2) — using a mapping tool that is typically treated as an authoritative visual record.
The risk wasn't lost on observers before Google's own statement. A BBC journalist posted sarcastically on Thursday, the same day of launch: "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," as reported by TechCrunch (E8).
What Did Google Say Officially, and What Is It Doing Now?
A day after launch, Google issued a new statement explaining the rollback. The company said, "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. 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," according to The Verge (E4).
TechCrunch's report of the same statement separates the two key admissions: Google acknowledged seeing "geospatial professionals using this feature for a range of useful purposes," while also seeing "people sharing screenshots of generated imagery that appear to violate our policies" (E10). Google added that it is "rolling back this feature in Google Earth while we work on implementing stronger guardrails" (E11) — meaning the feature is paused, not confirmed dead, pending new safeguards.
What Are Critics' Core Concerns About Misinformation?
Beyond the specific fabricated images van Ess produced, TechCrunch summarized the broader concern driving the backlash: "critics quickly pointed out that the tool could be used to create and spread misinformation" (E7). That concern centers on Google Earth's status as a trusted reference tool — the same trust Google itself cited in its rollback statement (E4, E10) — which critics argued made AI-generated fabrications inside the app especially risky if mistaken for authentic satellite imagery.
Timeline at a Glance
| Event | Timing | Source |
|---|
| Feature launches with Nano Banana 2 editing | Thursday | TechCrunch (E6) |
| Van Ess generates refugee/bomb-crater images | Thursday, post-launch | The Verge (E2) |
| BBC journalist posts sarcastic warning | Thursday | TechCrunch (E8) |
| Google's initial reply cites watermark + harmful-topic blocking | Thursday | The Verge (E3) |
| Van Ess fools Hive detection tool with generated video | After initial reply | The Verge (E5) |
| Google rolls back feature, cites policy-violating shares | One day after launch | The Verge (E4); TechCrunch (E9, E10, E11) |
What This Means
The sequence laid out across both outlets shows a gap between Google's stated safeguards and what testing actually found: the company's Thursday claim that watermarking and harmful-topic blocking were in place (E3) was directly undercut by van Ess fooling a detection tool with generated video the same feature produced (E5), and by his success generating images of refugees and a Gaza hospital bomb crater (E2). Google's own rollback statement acknowledges this tension without resolving it — crediting "geospatial professionals" for legitimate use while citing policy-violating screenshots as the reason for pulling the feature (E10), and promising "stronger guardrails" rather than a permanent cancellation (E11). The one-day span between launch and rollback (E6, E1, E9) suggests the gap between the feature's intended safeguards and its real-world use surfaced faster than Google anticipated.