On Thursday, Google added a new capability to Google Earth that let users apply Nano Banana 2, its AI image generator, to create images layered onto satellite maps. Google said the intention behind the feature was to enable creative uses of geographic content.
Shortly after the rollout, critics raised concerns that the prompt-based tool — which allowed almost any image to be superimposed onto real map views — could be misused to produce and spread misinformation. A BBC journalist sarcastically noted on Thursday that it was obvious how such a feature on one of the most relied-upon visual evidence sources for journalists and researchers could be abused online.
Only one day after the feature went live, Google removed it. In a statement, the company said, “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.”
“We’re rolling back this feature in Google Earth while we work on implementing stronger guardrails,” Google added.
It is worth noting that in the age of AI, most imagery online is vulnerable to manipulation. Producing semi-credible disinformation no longer requires advanced Photoshop skills; access to an AI image generator is often sufficient — tools that Google and many other AI firms offer.
Why this matters
Platforms like Google Earth are particularly sensitive to the authenticity of visual sources because journalists, researchers and decision-makers rely on them to document local events and geographic phenomena. The quick rollback underscores the risks associated with deploying AI features without robust safeguards and the need for developers to design protections before broad releases.
