Google is making its Project Genie world model more connected to real locations by adding Street View support, allowing users to create interactive AI-generated environments based on actual places.
Google DeepMind is connecting Street View to Project Genie, its general-purpose world model that can generate diverse, interactive environments. Google also confirmed in its official blog that the update lets Genie use real-world imagery from Street View so users can explore familiar places or reimagine them with creative styles.
Street View becomes part of Google’s world model
The new feature was introduced during Google I/O 2026, where Google showed how Genie can move beyond text prompts and images by grounding AI-generated worlds in real streets.
TechCrunch reported that the goal is to make Street View more immersive and interactive, letting people simulate streets and nearby surroundings instead of only viewing static map imagery. The article gave examples such as changing the weather or imagining a place in a disaster-like “Day After Tomorrow” scenario.
Google’s blog explained that users can tap a Maps pin to choose a place in the United States, select a style such as “Desert Sands” or “Stone Age,” and describe a character like an animal, comic book hero, or claymation monster. Genie then creates an imaginative world tied to Street View’s real-world imagery.
Google wants Genie to help humans, agents, and robots
The update is not only for playful exploration. Google is also positioning Genie as a tool for AI agents and robotics training.
Jack Parker-Holder, a research scientist on Google DeepMind’s open-endedness team, shared that Genie is powerful for both “agent” and robotics use cases, as well as for people who simply want to play with the technology. He gave the example of a robot deployed in London, where Genie could simulate rare sunny conditions so the robot would not be surprised when sunlight reflects off Victorian housing.
Connecting Genie with Street View can provide a virtual environment where AI agents or robots can navigate and interact with real-world complexity. It also noted that Genie has already become a foundational research tool for agents learning and reasoning in complex virtual settings.
Street View gives Genie a huge real-world base
Street View gives Google a major advantage because it has been collecting real-world imagery for years.
Google has collected Street View data for 20 years using camera-equipped cars and people with tracker backpacks. The company has gathered more than 280 billion images across 110 countries and seven continents, giving Genie a large base of real-world visual information.
Google said the new capability is powered by Maps Imagery Grounding, the same technology developers use to create AI visuals with Street View. The company gave examples such as placing the Golden Gate Bridge under the sea through an “Ocean World” style or exploring the Fort Worth Stockyards in Texas with a black-and-white film look.
The feature is still experimental
The new Street View capability is available now for places in the United States, with Google planning to expand it to more locations over time.
Google said Project Genie, including the Street View feature, is gradually rolling out to eligible Google AI Ultra subscribers globally who are at least 18 years old. The company also described Genie as an experimental research prototype in Google Labs and said it is still working to make details sharper and more accurate.
In demos shown by Google, the results were recognizable but still closer to video game quality than photorealistic simulation. The models are not yet physics-aware, meaning they do not fully understand cause and effect.
A step toward AI-generated virtual worlds
The bigger significance of the update is that Google is trying to connect generative AI with real-world geography.
Jonathan Herbert, a Google Maps director, said that Genie cannot yet create a fully faithful reconstruction of a street. But he said the breakthrough is spatial continuity, where the AI can remember what is behind the user after they turn around and continue building the environment from there.
For users, the feature could make Street View feel more like an explorable game or simulation. For Google, it shows how Maps data, Street View imagery, and AI world models could eventually support entertainment, education, robotics, and autonomous systems that need safer ways to practice in realistic environments.
