Overview
- Gemini’s update, announced Thursday, will reach AI Plus, Pro, and Ultra subscribers in the U.S. over the next few days, with desktop Chrome and more regions to follow.
- The assistant now links its Personal Intelligence system to the Nano Banana 2 image model so short prompts produce images that match your tastes.
- If you opt in to connect Google Photos, Gemini uses your labels to identify people or pets and can make pictures of “me and my family” without manual uploads.
- A Sources button shows which photo guided the result, and a plus icon lets you choose a different reference or refine the image.
- Google says private Photos will not directly train its models, though it learns from prompts and responses in the app, a stance some outlets flagged for trust and privacy concerns.