Google Photos to auto-build a digital wardrobe with virtual try-on

TL;DR: Google Photos will roll out a new wardrobe feature this summer, using AI to scan users’ photo libraries and automatically catalogue clothing items into a categorised digital closet. The feature will support filtering by category (jewellery, tops, bottoms), outfit moodboards for occasions and trips, and a virtual try-on preview that overlays selected pieces. Android first, iOS to follow.

The feature is the latest in a steady stream of generative-AI capabilities Google has added to Photos over the past two years, alongside Magic Editor, Best Take and Memories albums. The wardrobe takes a familiar pattern — automatic clustering by content — and applies it specifically to garments, with the practical promise that users can rediscover items they own but have forgotten, mix and match across past photos, and visualise outfit combinations before getting dressed.

What’s new and what’s quietly significant

The virtual try-on element is the more substantive AI workload. Rather than simply organising existing photos, it generates a synthetic preview — applying selected clothing items to the user’s likeness based on their existing photo library. Google has not detailed which model powers this, but the capability sits within the lineage of try-on tools that emerged in Google Shopping experiments last year, and it implies that Google is comfortable running on-account image generation for consumer accounts at this kind of scale.

For UK retailers, the wardrobe is also a quiet shift in the customer-discovery surface. ASOS this week detailed its own “AI Stylist” conversational shopping experience built on Microsoft’s stack. Google Photos is now offering a comparable styling layer that sits outside any retailer’s app, draws on the user’s actual closet rather than a catalogue, and feeds whatever inspiration it surfaces back into Google’s own search and shopping graph.

Looking forward

UK consumers should expect the feature this summer, with the usual default-on questions about whether clothing-detection metadata is processed on-device or server-side, and how it interacts with Google Photos’ existing face-grouping consent flow. The ICO has previously taken interest in automatic personal-attribute classification at scale; clothing inventories are not biometric but represent a fresh category of inferred personal data.

For UK fashion retailers, the practical question is whether to integrate with Google Photos’ wardrobe surface as it develops, or invest in proprietary AI styling tools as ASOS has chosen to do. The differentiator will likely be data: a closet-aware competitor to a retailer’s own styling tool is a meaningful change in the customer relationship.