Google's New AI Image Tool Reveals How We Really Use It
Most people use AI image tools for fun, not work—here's what that means for you.
Using Urania, a Google-developed framework with rigorous differential privacy guarantees, the analysis examined a subsample of 2.1 million early-2026 conversations that used Gemini’s Nano Banana models—without ever looking at a single image or individual prompt. It grouped them into roughly 800 clusters across five categories: Social & Viral Sharing, Brand & Marketing Collateral, Educational & Technical Information Media, Product Visualizations & E-commerce, and Architectural & Spatial Designs.
The clusters were compared on 21 features of use, measuring prompts’ editing techniques, artistic style requests, composition, and chatbot interactions. This provides a privacy-preserving baseline for how people use image generation and editing tools across different use cases, including growing and emergent use cases that were not possible before Nano Banana.
- Most people use Google's AI image tool for social media and fun, not just work.
- The study analyzed 2.1 million conversations without seeing any private images or prompts.
- New uses like interactive product visuals and architectural design are emerging, making advanced editing accessible to everyone.
Why It Matters
AI image tools are becoming mainstream for everyday creativity, changing how we communicate and shop online.