Google Photos New Feature: AI scans old photos, Wardrobe creates your virtual fitting room

Google Photos is about to launch the “Wardrobe” feature, which automatically scans users’ albums, creates a digital wardrobe, and supports virtual try-ons; the core is the “Nano Banana” model, responsible for recognizing clothing items, generating digital avatars, and rendering try-on effects; the feature will first be available on Android devices (summer 2026), with iOS support to follow later.
(Background recap: Nano Banana 2 is here! Supporting 4K enhanced character consistency, Gemini raw images faster and more accurately)
(Additional background: a16z’s latest AI Top 100 applications list: ChatGPT remains at the top, platform brand positioning battle officially begins)

Table of Contents

Toggle

• How AI wardrobe works
• How this differs from last year’s try-on feature
• Mobile albums are becoming another entry point

A memorable scene from the 1995 film “Sense and Sensibility”: the heroine stands in front of a screen, and the computer automatically fetches clothes from the wardrobe, matching a look for the day in real-time. Today, that scene is gradually becoming a reality with AI assistance.

Google announced that Google Photos is officially launching the “Wardrobe” feature. The system will proactively scan users’ albums, recognize clothes worn in past photos, automatically organize them into categories like tops, bottoms, skirts, dresses, shoes, etc., and create an interactive digital wardrobe.

How AI wardrobe works

The core of the feature is the “Nano Banana” model. The system will generate a digital avatar of you, then render your selected clothing combinations onto that avatar to complete a “virtual try-on.”

The entire process runs on the device. Users don’t need to upload any additional data; the system processes existing albums locally. Favorite combinations can also be saved as a “moodboard” or shared with friends.

However, it’s important to note that Nano Banana is responsible for recognition and rendering, not for “advising you on what to wear today.” It is a tool, not a stylist.

How this differs from last year’s try-on feature

In 2025, Google introduced an AI try-on feature in Search, but that version was designed for: when you search for a product on Google, you want to see how it looks on you, and the system lets you try it on before deciding whether to buy. The core logic was “shopping assistance.”

This time, the Wardrobe feature’s logic is entirely different. The object of application shifts from “products you’re considering buying” to “clothes you already own,” moving from a consumer behavior perspective to daily management. Your albums are no longer just a memory repository but become an operable clothing database.

Some Android devices (like Motorola Razr) are already testing this, with the official release expected before summer 2026. The iOS version will follow afterward.

Mobile albums are becoming another entry point

In recent years, Google has continued integrating AI capabilities into its consumer apps: Gemini into Gmail, Docs, Maps; AI-generated summaries appearing in Search results; Photos app with Magic Eraser to remove clutter, Photo Unblur to restore blurry photos.

Wardrobe is a continuation of this trend, but the scene is more private this time: your clothing history.

However, a question worth pondering is: when AI systems can automatically build a comprehensive profile of your clothing preferences from your photos over the past few years, where are the ownership and usage boundaries of this data? Google has not clarified whether the Wardrobe feature will affect ad targeting, nor whether the digital wardrobe data will be used for model training.