Back to today’s briefing

Story perspectives

Google Earth AI Tool Pulled Over War-Zone Misinformation Fears

8/2/2026

35 8 Full Breakdown

Summary

Google Pulls AI

On July 30 2026, Google added the Nano Banana 2 AI image-generation tool to Google Earth’s web version.
The feature, limited to earth.google.com and absent from iOS and Android apps, can create historic scenes, infographics and lot visualizations anchored to the viewport’s satellite and 3D data.
After policy-violating screenshots appeared, Google rolled back the tool, watermarked its outputs and moved it to a separate subscription tier.

Summary

Researchers expose AI

Researchers Eliot Higgins and Henk van Ess showed Nano Banana 2 can create realistic war-zone images, sparking misinformation worries.
BBC Verify found the SynthID watermark only partially applied and that some AI-detection tools missed the fakes.
Yale’s Nathaniel Raymond warned that the tool opens floodgates for spoofing satellite data and that TerraWatch Space CEO Aravind Ravichandran cautions it could erode journalists’ trust in Google Earth.

History

Historical Context

In 2018 BuzzFeed released a deep-fake video of former President Barack Obama, the first public exposure of AI-generated speech and facial manipulation. The method reappeared in 2026 when Trump’s Truth Social post showed AI-generated apes of the Obamas, leading regulators to require watermarking under the 2022 AI Bill of Rights. Google’s Nano Banana 2 tool now creates synthetic satellite scenes, noted by Eliot Higgins and Henk van Ess.

1 of 3

Google Pulls AI

On July 30 2026, Google added the Nano Banana 2 AI image-generation tool to Google Earth’s web version.
The feature, limited to earth.google.com and absent from iOS and Android apps, can create historic scenes, infographics and lot visualizations anchored to the viewport’s satellite and 3D data.
After policy-violating screenshots appeared, Google rolled back the tool, watermarked its outputs and moved it to a separate subscription tier.

1 / 3