Gemini AI Uses Your Photos for Smarter Images

Google has enhanced its Gemini AI with a "personal intelligence" feature that now connects directly to Google Photos. This integration, available to AI subscribers, allows the AI image generation model to access a user's photo library and associated labels. The goal is to simplify prompt creation and improve the accuracy of generated images by providing contextual information directly from the user's existing photos.

Users can now reference groups like "my family" or "my dog" in prompts, and Gemini will leverage their Google Photos content to identify and create the desired image. For instance, a prompt for "a claymation image of me and my family enjoying our favorite activity" would utilize photo labels to identify family members and image content to infer favorite activities. This reduces the need for extensive textual descriptions in prompts.

Google assures users that while personal data from inputs and outputs helps improve AI products, images from the user's library are not used for training the AI models. Personal intelligence is opt-in and off by default, with access currently limited to paid Google AI plans. However, similar features in the past have expanded to broader user bases over time, and Gemini is noted for encouraging users to enable this connectivity.