What happened to Google Earth's AI image tool?
Google Earth briefly allowed users to create AI-modified versions of satellite imagery, but quickly reversed its decision due to concerns over misinformation and disinformation. The feature, which used the Nano Banana 2 AI image generator, made it easy for anyone to create fake scenarios overlaid on real locations. Researchers and investigators sounded the alarm, showcasing the potential for abuse.
Google Earth's introduction of an AI tool for modifying satellite imagery was met with instant backlash from researchers and investigators - and it's easy to see why. This feature, powered by the Nano Banana 2 AI image generator, let users create custom images using Google Earth's satellite, aerial, and 3D imagery. Google pitched it as a way for users to show off completed real estate projects or reimagine historical sites, but others were quick to point out the potential for trouble.
The potential for misuse was pretty obvious from the start, as users began sharing examples of AI-generated images that showed just how easily misinformation and disinformation could spread. Henk van Ess, an independent investigator, demonstrated just how simple it was to create fake scenarios - like planting a nuclear plant in Iran or depicting refugees near the Mexican border. These images, overlaid on Google's super-realistic satellite imagery, were seriously convincing.
The concerns about misinformation and disinformation aren't unfounded, it's worth noting. With the rise of generative AI tools, it's become way too easy to create fake images and videos that can be used to deceive or manipulate. And Google Earth, a tool often used to verify images and videos of locations, was particularly vulnerable to abuse. By letting users modify satellite imagery with AI, Google inadvertently created a powerful tool for spreading misinformation - which is why it was so important for them to roll back the feature.
In response to the outcry, Google quickly backpedaled, announcing on X that it was ditching the AI tool from Google Earth. It was a smart move, really, as it prevents the potential for widespread misuse. But it also highlights the need for tech companies to think really carefully about the implications of their actions and to prioritize the integrity of their platforms. As AI tech continues to evolve, it's crucial that companies like Google take steps to prevent the spread of misinformation and disinformation - it's a big responsibility, but someone's gotta do it.
Source: Ars Technica
Related posts
Mistral Large 4 Arrives on Hugging Face With Cybersecurity Focus
Mistral Large 4, a 1-trillion-parameter multimodal mixture-of-experts model with 49 billion active parameters and a 1-million-token context window, launched on Hugging Face October 29, 2026. It scores 82 percent on the Artificial Analysis Cyber Index for reproducing and patching vulnerabilities, 93 percent on Cybench, and 93.3 percent on Lakera's prompt-injection benchmark. Released under the Mistral Ecosystem Open-Weight License, it supports day-zero inference via transformers, vLLM,
NVIDIA Releases Nemotron 3.5 Transcribe for Promptable Multilingual
NVIDIA released Nemotron 3.5 Transcribe, an open-weight model on Hugging Face and ModelScope (October 28, 2026) that transcribes 40 languages, translates 31 to/from English, and attributes speakers in 13 languages-all via natural language prompts. The 30-billion-parameter model activates only 4 billion per token, topping open-weight benchmarks with 5.33% English WER and 18.26% diarization error rate across 21 languages. Granary v2 fine-tuning toolkit accompanies it.
Google DeepMind Unveils Gemini 4 Carbon Model
Google DeepMind unveiled Gemini 4 Carbon on October 28, 2026, claiming it as their most capable model yet with state-of-the-art performance in cybersecurity and long-horizon reasoning. The company says Carbon tops the updated v5 benchmark suite, though the specific test wasn't fully named. Technical details, parameter counts, and architectural changes remain undisclosed, and no timeline was given for API, consumer app, or NotebookLM availability. Independent evaluations and public acce