New AI Feature in Google Earth
Recently, Google introduced a new feature in Google Earth. It allows users to create images using artificial intelligence. All they need to do is pick a spot on the map and type in a few words. Suddenly, they can generate visuals that show anything from buildings burning to destruction in famous places. This feature came with major concerns because there were no proper checks in place. The potential for spreading false information became clear.
Google integrated its AI generator, Nano Banana, into the service. The result? People could now make images of offices on fire or massive damage to places like San Francisco or the Eiffel Tower. These images looked real, and they were created in seconds. That kind of power is dangerous. It gives bad actors a simple way to spread misleading content. The visuals are not just convincing, but also very easy to produce.
The company rolled out the AI feature in Google Earth and quickly faced pushback. Online users began sharing examples of disturbing images. The backlash forced Google to respond. At first, they told people to check images they found suspicious. But as the fake images spread more, the company decided to take action. It announced that it would roll back the feature. The plan was to add stronger safeguards before re-releasing it.
Experts and reporters were quick to point out the risks. One writer on LinkedIn warned that the tool could be used to create images that look exactly like real drone strikes. These look-a-likes are hard to tell apart from real ones, especially in places like Ukraine, Russia, or the Middle East. The ease with which such images can be generated and shared is alarming. It makes it harder to know what's real and what's fake.
Past AI Missteps
This isn’t the first time Google has had problems with AI features. In the past, the company released tools without enough safety measures. One example was an AI shopping tool that let users create images of clothes on their own bodies. But it had a dark side. People could make inappropriate images of celebrities or even minors. That caused serious ethical issues.
In 2024, Google Gemini faced criticism for another problem. It generated images of racially diverse groups of Nazis but not white soldiers. That raised concerns about bias in AI tools. Another issue came with Google's chatbot-powered search. It gave dangerous health advice, like telling people to eat rocks. After a lot of public criticism, Google made changes. Yet even two years later, its AI search still has problems. It still gives inconsistent answers.
Google isn’t the only company to make mistakes with AI. Elon Musk's Grok app once generated more than a million sexualized images of children. It happened in just a few days. Other companies, like OpenAI with ChatGPT, have also had embarrassing errors. These examples show that AI tools need better control and oversight. Otherwise, they can cause real harm.
Google Earth has long been an important tool for researchers, journalists, and even government officials. It helps them study conflicts, natural disasters, and more. The satellite images can show things like crop failures, deforestation, and mass graves. But now, with the new AI feature, the service could become a tool for spreading lies. That's a big shift in what it can be used for.
One of the most worrying parts is the speed at which users can create fake content. For example, someone could generate an image of a Trump Hotel destroyed in the Gaza Strip. It's done in seconds. That makes it easy for misinformation to spread fast. It's especially concerning because users don't even need to search for images on other AI tools. The process is all built into one simple tool.
Google tried to defend itself by saying that every AI-generated image from the tool includes a digital watermark. They said people can use their Gemini app or Lens in Search to check. But critics argue that's not enough. The tool is already being used in ways that are hard to control. Some people are using it to make images that look like real events in war zones. That’s a big red flag.
Another strange thing is that some AI features in Google already have safety checks. For example, the company's Nano Banana tool refuses to edit images to show terrorist attacks, like 9/11. But when the same idea was introduced into Google Earth, there were no such restrictions. That inconsistency shows that Google may not have considered all the risks before launching the new feature.
Previous testing of Google’s SynthID system has shown that Google’s digital watermark can persist despite substantial image edits and data loss. However, it limits image checks to about 10 per day. When Ars prompted Google’s Gemini app to run a SynthID check on several AI-modified images generated within Google Earth—before Google removed the Google Earth feature—Gemini responded that it had 'detected digital watermarking indicating that part or all of this image was generated or edited using Google AI tools.' But Ess questioned why Google’s policy on preventing the creation of AI-generated images on harmful topics had not stopped him from depicting scenes such as a hospital with a bomb crater in Gaza. He also expressed skepticism about primarily relying on Google’s SynthID watermarking system and Google’s AI tools for verifying whether an image is AI-generated. 'Fakes do not travel as clean files with their credentials intact,' Ess wrote. 'They travel as screen recordings, re-encodes, screenshots of screenshots, filmed off somebody’s phone in a hurry.' Indeed, Google’s SynthID was unable to provide any image verification when Ars used a smartphone camera to take a photo of an AI-altered image made using the Google Earth feature.
By introducing an AI image generator as a Google Earth feature, Google risked undermining public trust in Google Earth as a reliable reference for providing authentic imagery of the real world, Ess said. It could have also handed an additional excuse to government officials or anyone who wants to deny genuine satellite photos by claiming they were modified or generated entirely by AI tools. Giving anyone access to an easy tool to create AI-modified satellite imagery could have also made it even harder for journalists and researchers checking misinformation and disinformation. Ess recommended using multiple sources to verify satellite imagery and relying on constantly updated satellite imagery sources such as Europe’s Copernicus Sentinel-2.
Google may have alleviated immediate concerns by deciding to quickly remove the AI tool within Google Earth—but its statement suggested that the company could eventually bring the feature back in the future. '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,' according to the Google statement. 'So we’re rolling back this feature in Google Earth while we work on implementing stronger guardrails.'
Looking at past events, it seems clear that Google could have known better. Its AI tools have caused problems before. From inappropriate image generation to giving dangerous health advice, the company has a history of launching features without enough testing. That's why users and experts are so worried about the new AI feature in Google Earth.
As Google works on improving the feature, it needs to make sure the changes are strong enough. The goal should be to stop fake images from being created and shared. The digital watermark is a start, but it's not a full solution. Until the company can control the tool better, it's hard to trust that the service won't be used for harm. For now, the feature is on hold. But the lessons from this rollout are clear: AI needs to be used with care and responsibility.