What happened
On Thursday, July 31, Google rolled out a feature letting Google Earth users generate AI imagery with its Nano Banana 2 image model and overlay it directly onto real satellite maps, pitched as a tool for creative geographic experimentation. Within a day, Google pulled the feature. In a statement, the company said it had "seen people sharing screenshots of generated imagery that appear to violate our policies" and was "rolling back this feature in Google Earth while we work on implementing stronger guardrails." Critics moved fast: a BBC piece flagged the risk almost immediately, and social media users pointed out the obvious problem, that Google Earth is widely treated as a reliable, neutral source of visual evidence by journalists, researchers, and courts, and a prompt-based tool that can superimpose any image onto real coordinates undermines exactly that credibility.
Why it matters for your business
This is a fast, real-world example of how quickly an AI feature can go from launch to rollback when a company underestimates misuse, and it is a useful cautionary tale if your business is adding AI generation tools, chatbots, or image features to your own website or app. Google has enormous trust and safety resources and still had to pull a feature within 24 hours. If you are building or buying AI tools for customer-facing use, plan for a moderation and rollback process before launch, not after a problem surfaces. It is also a reminder to your team: content generated by AI tools, including maps and official-looking imagery, should not be treated as verified fact.