ChatGPT adds virtual try-on, pushing AI deeper into the shopping flow
OpenAI's new try-on and favorites features insert an AI layer between shoppers and retailers' own product pages, a year after Google shipped something comparable.
Why it's worth posting
OpenAI's global launch of a virtual try-on feature and a favorites function, both built on the ChatGPT Images 2.5 model, reads as an incremental product update but carries a structural consequence worth covering. By letting a shopper upload a selfie or full-body photo and preview clothing inside ChatGPT, OpenAI inserts an AI intermediary between the shopper and the retailer's product pages, capturing the discovery moment. That forces a real decision onto fashion e-commerce platforms: feed product data into ChatGPT's ecosystem to extend reach, or risk ceding that moment to it. Google launched a comparable try-on feature a year earlier, so the competitive window is not opening now, it is already narrowing. The favorites-and-library function signals OpenAI is building persistent shopping behavior inside the app, which is the institutional shift a creator covering commerce or AI product strategy can track.
The visible change is the try-on feature; the harder question is what the favorites and library function implies. Saving products and try-on images into a persistent library is how a platform builds durable shopping behavior inside its own walls, rather than acting as a one-off tool. That is the piece worth watching for anyone covering retail technology.
The competitive clock matters. Google shipped a comparable try-on feature a year before this announcement and has separately announced an AI digital closet, which complicates any framing of OpenAI's move as breaking new ground. The quality advantages OpenAI cites, more natural lighting and richer textures, come solely from the company, so a creator should treat them as claims pending independent testing.
There is also a track-record caveat. OpenAI previously offered an instant checkout feature that did not perform well, and this launch does not close the loop from try-on to purchase. The gap between visualizing an item and buying it remains open, which is both a limitation of the current feature and the obvious next battleground.
Angles to take
Follow the structural consequence: with an AI layer now sitting between shoppers and retailers' product pages, fashion e-commerce platforms face a real choice between feeding inventory into ChatGPT or losing the discovery moment to it.
Write this post →Frame it as a workflow collapse: conversation, product discovery, visual try-on, and saving favorites now live in one general-purpose assistant, a flow that previously spanned several apps.
Write this post →Take the skeptic's line: Google shipped a comparable try-on feature a year earlier, the quality claims come only from OpenAI, and the company's prior instant-checkout feature flopped, so question how transformative this entry actually is.
Write this post →