Google's Gemini can now reference photos from a user's Google Photos library when generating personalized AI images. The feature uses machine learning to analyze selected photos and incorporate visual context. Google clarifies that photos are not retained for training, though user inputs and outputs may improve other products; the feature is opt-in, off by default, and available only on paid plans.
Products
Gemini can now create personalized AI images by digging around in Google Photos
Gemini now generates personalized images by analyzing users' Google Photos library as visual reference, available opt-in on paid tiers without retaining source photos for model training.
Thursday, April 16, 2026 12:00 PM UTC2 MIN READSOURCE: Ars TechnicaBY sys://pipeline
Tags
products
/// RELATED
Research5d ago
Liquid Neural Network Models for Natural Gas Spot Price Time-Series Forecasting
Researchers apply liquid neural network architecture to natural gas price forecasting, exploring a novel deep learning approach for commodity market time-series prediction.
Strategy5d ago
OpenAI is ‘strongly positioned,’ says Wedbush’s Dan Ives
OpenAI missed targets (900M vs 1B weekly actives, revenue goals) triggering AI stock selloffs, but Wedbush analysts dismiss the panic as overblown given OpenAI's $600B compute plan and strong market position.