Google Withdraws Earth AI Image Tool Over Fake Satellite Imagery Concerns

Google said on Friday it is withdrawing a new artificial intelligence image-generation feature from Google Earth after users created altered satellite images that appeared to violate its policies.

The feature, introduced on Thursday, allowed users to generate photorealistic aerial scenes using text prompts. Concerns quickly emerged that the fabricated images could be mistaken for genuine satellite imagery.

Google said in a post on X that it was rolling back the feature to introduce stronger safeguards. The company stressed that none of the AI-generated images had appeared on the main Google Earth platform.

Dutch intelligence researcher Henk van Ess raised concerns about the tool on Thursday, demonstrating how it could be used to create a fabricated nuclear facility in Iran.

Van Ess warned that increasingly realistic AI-generated landscapes could make it harder for journalists, intelligence researchers and fact-checkers to distinguish genuine satellite images from manipulated material.

A Google spokesperson referred to the company’s statement on X when contacted for further comment.

Google has invested billions of dollars in artificial intelligence and has been adding generative AI features across products including Search, Workspace, Maps and Google Earth.

The company is seeking to show investors that its heavy spending on AI infrastructure can produce new revenue and increase user engagement. Alphabet shares recently came under pressure after the company outlined a $205 billion spending plan that raised concerns about the cost of its AI expansion.

The withdrawal of the Google Earth feature illustrates the risks technology companies face when releasing increasingly realistic generative AI tools that can be used to produce misleading content.