Google's Gemini AI Now Crafts Images From Your Photo Library

Google is introducing a more intimate dimension to AI image generation. Its Gemini assistant can now produce pictures informed by a user's personal Google Photos library, provided they have opted into the broader 'Personal Intelligence' feature. This integration allows the AI to reference a user's photos to shape its creations. For example, a prompt such as "Design my dream house" could yield an image reflecting architectural styles or color palettes the user has previously captured and saved.

According to Google, the system uses labels within Photos to identify subjects like the user, their friends, or family, then employs its Nano Banana 2 image model to generate the new, personalized content. The company emphasizes a distinction in its data handling: while it will not directly train its core AI models on private photos, it does analyze interactions with the feature itself, including specific prompts and the model's corresponding outputs. This new capability is beginning a staged release for paying Gemini subscribers in the United States, with plans to expand to desktop Chrome users and other regions shortly. The move signals a clear industry shift toward AI tools that leverage personal context, raising both possibilities for customization and ongoing questions about data privacy boundaries.

Source: The Verge