Overview
- Google said Thursday, April 16, that the Gemini app is rolling out personalized image creation over the next few days to eligible AI Plus, Pro, and Ultra subscribers in the U.S., with support for Chrome on desktop coming soon.
- Personal Intelligence now feeds Nano Banana 2 with account-level context from connected Google apps, so short prompts like “Design my dream house” can reflect a user’s tastes without long descriptions or manual uploads.
- If users connect Google Photos, Gemini can use their existing people and pet labels to auto-select reference images of specific individuals, and a Sources button and a “+” icon let users see, swap, or refine those references.
- Google says the app does not directly train models on private Photos libraries and instead learns from limited information such as prompts and model responses, and all connections remain opt-in and can be changed in settings.
- Coverage notes rising privacy scrutiny, with some outlets flagging Google’s careful wording on training and citing a proposed class-action in San Jose over Gemini activation, even as Google positions the feature as a time-saver for everyday creative tasks.