Now, shoppers using Google’s mobile website or apps can refine their searches with tools designed to help them narrow down exactly what kind of jacket or other item they’re looking for. The Style Recommendations feature is rolling out now for signed-in shoppers in the US, and it lets users rate items in their searches so it can pick up on their preferences, kind of like Tinder or Netflix ratings but with shirts.

After they’ve found clothes they like, shoppers can specify the brands they like, which will instantly bring up more apparel from those selections. 

Google previously tested AI tools that help people find more of the items they like. In November, users on the company’s Search Generative Experience tested out an image generation feature that puts together jackets and other clothing based on prompts. People can then find similar items based on the AI image. The company also tested a virtual try-on so shoppers can see how dresses from brands look on them. Both features are now live on Google Shopping. 

Google joins other e-commerce platforms in bringing generative AI tools to shopping. Amazon’s shopping assistant, Rufus, answers customer queries about products. While the chatbot began making up jokes about Jeff Bezos, Amazon rolled it out to select users. The company also offers AI-powered sizing recommendations.

However, AI-powered shopping features are still new and can lead to questionable items. The Verge’s Mia Sato said the AI shopping experience proved confusing and yielded clothes that weren’t prototyped to fit other humans. 

Correction, March 27th, 1:47 PM ET: An earlier version of this story claimed Google is adding more AI in shopping. In fact, the new style recommendations and refining search do not use generative AI technology like the image generation and features that were previously added, and the headline and text now reflect this. We regret the error.