By The Chronicle Staff

Google has temporarily disabled a new artificial intelligence feature in Google Earth after users quickly demonstrated how it could be used to create realistic—but entirely fake—satellite images of real-world locations.

The feature, introduced as an experimental tool, allowed users to edit satellite imagery using simple text prompts. Within hours of its release, social media filled with fabricated images showing events that never happened, prompting Google to roll back the feature less than a day after launch.

The AI tool was powered by Google’s latest image-generation technology and was designed to let users creatively modify aerial and satellite views. While some geospatial professionals used it for harmless demonstrations and planning concepts, others immediately began generating convincing images of natural disasters, military installations and conflict zones that did not exist.

Among the fabricated scenes shared online were fake flood damage, fictional refugee camps, imaginary nuclear facilities and simulated attacks on recognizable locations. Although the images carried AI watermarks and were never added to Google’s public map data, experts warned they could easily be downloaded and shared without context across social media.

Google responded swiftly.

The company announced it was “rolling back” the feature while it develops stronger safeguards to prevent misuse. In a statement, Google emphasized that maintaining trust in Google Earth is critical and said the AI-generated images were always separate from the platform’s authentic satellite imagery.

The incident highlights one of the biggest challenges facing artificial intelligence today: balancing innovation with the potential for misinformation.

Google Earth has long been considered a trusted source for viewing satellite imagery. Journalists, researchers, emergency responders and investigators frequently rely on the platform to verify locations and analyze real-world events.

Introducing AI-generated imagery — even when clearly labeled — raises concerns that fabricated images could be mistaken for genuine evidence once they leave the platform.

Experts in digital forensics say the timing is particularly significant.

As AI-generated images become increasingly realistic, distinguishing fact from fiction is becoming more difficult. While many AI systems include invisible watermarks or metadata, those protections can be lost when images are cropped, compressed or reposted on other platforms.

The result is an online environment where viewers may find it harder to know whether an image documents a real event or was created in seconds by artificial intelligence.

Google’s decision also reflects a broader trend across the technology industry. Several companies have recently scaled back or modified AI features after discovering they could be exploited in unexpected ways, especially when realistic images involve public figures, disasters or sensitive locations.

For consumers, the episode serves as another reminder to approach striking images online with healthy skepticism.

Whether a dramatic satellite image appears on social media, in a group text or on a

website, experts recommend looking for confirmation from multiple reputable news organizations before accepting it as authentic.

Artificial intelligence is becoming an increasingly powerful creative tool, but as Google’s brief Google Earth experiment demonstrated, the same technology that can imagine new possibilities can also create convincing versions of events that never happened.■