Global technology firm Google identified five distinct categories of AI image generation usage patterns and user behaviors based on an extensive analysis of user interactions, the company reported.
The research, which utilized Google’s Urania framework, analyzed 2.1 million conversations from its Gemini Nano Banana platform to understand how users engage with AI image creation tools, moving beyond common perceptions of social media-centric use.
The five identified categories include Social & Viral, Brand & Marketing, Product Visualization & E-Commerce, Architectural & Spatial Design, and Educational & Technical Information, according to Google.
Analysis focused on features such as artistic styles, image editing capabilities, image composition, interaction characteristics, and multi-select options within user sessions.
Memes and humor were found to be the least prevalent features within the social and viral categories, Google stated. The inclusion of people in generated images was the most common feature, appearing in nearly 80 percent of instances within this category.
Commercial applications showed significant depth, with marketing usage frequently involving detailed composition controls. Many instances involved original or fictional brands, suggesting that users are developing new business concepts through AI image generation, Google reported.
The findings indicate a broader and more commercially driven application of AI image generation tools than previously understood, extending into specialized fields like product development and architectural planning.
Google conducted the research without compromising user privacy, focusing on aggregated and anonymized data to discern these usage patterns.
Source: Search Engine Journal
