Google rolled out Personal Intelligence image generation in the Gemini app, using its Nano Banana 2 model to pull from an opted-in user's Google Photos library and generate personalized images without manual uploads or detailed prompts. Users can ask things like "design my dream house" and the system uses photo labels to incorporate likenesses of the user, friends, and pets; a Sources view shows which photos shaped the output. The feature is opt-in and Google states it does not directly train models on private Photos libraries, though limited prompt and response data is used. Rollout is underway for Google AI Plus, Pro, and Ultra subscribers in the US, with Chrome desktop and broader audiences to follow.