Google launches AI ‘agents' to pick and pay for your everyday online shopping

Irish Independent

20-05-2025

The tech giant has unveiled a new feature at its 'I/O' conference that will let an AI 'agent' suggest, pick and even pay for things on your behalf.
The new AI shopping feature will use Google's access to 50 billion product listings, from small boutiques to large chains, to incorporate details like reviews, prices, colour options and availability.
'Behind the scenes, we'll add the item to your cart on the merchant's site and securely complete the checkout on your behalf,' the company says.
'When you've made up your mind, our new agentic checkout will help you easily buy at a price that fits your budget.'
Google says that the information will be updated constantly, with two billion product listings updated every hour. The feature is initially being launched in the US.
The tech giant also unveiled a new feature that will allow people to virtually try on billions of clothes listings from online shops by uploading a photo of themselves.
'When you're shopping for shirts, pants, skirts and dresses on Google, simply tap the 'try it on' icon on product listings,' the company says.
'From there you can upload a full-length photo of yourself and within moments, you'll see how that wedding-season maxi dress or playful shirt for your next vacation looks on you.'
Google says that the feature is powered by a new custom image generation model for fashion, which understands the human body and nuances of clothing, like how different materials fold, stretch and drape on different bodies.
'It preserves these subtleties when applied to poses in your photos,' the company says.
The "try on" experiment is rolling out in Search Labs in the US today.
The new online features were announced as part of a slew of new AI enhancements and services at Google's annual I/O conference in the US.
The tech giant, which is in an AI arms race against Microsoft-backed OpenAI, has deepened the integration of AI into its globally-dominant Google search engine with a new 'AI Mode' available in the US.
Google is also giving real-time speech translation a reinvigorated shot with a new Al-powered feature that instantly translates spoken languages between participants 'in real-time, preserving the quality, nuance, and personality in their voices, so you can communicate naturally with anyone, regardless of their native languages'.
Separately, Google will make its high-end Gemini 2.5 Flash and Pro tiers 'generally available' soon, the company says. It will also launch Deep Think, an enhanced reasoning mode in Gemini 2.5 Pro, with higher performance on complex maths and coding tasks.
The company is also introducing a new top-tier AI subscription plan, Google Al Ultra, 'with the highest usage limits and access to our most capable models and premium features'. This new plan will initially roll out in the US for $250 (€224) per month,
Generative media models with Imagen 4 and Veo 3, its video and image generation models with native audio generation, will also be introduced, the company said.
And there's expanded access to Lyria 2, 'giving musicians more tools to create music', while a new Al filmmaking tool called 'Flow' is also being launched.
Google is also enhancing its AI-based coding capabilities via a public beta of 'Jules', an asynchronous agentic coding assistant that integrates directly with a coder's existing repositories and handles tasks on the user's behalf.
Meanwhile, camera and screenshare capabilities will be available to everyone on Android and iOS.
"More intelligence is available, for everyone, everywhere,' said Sundar Pichai, CEO of Google and (parent firm) Alphabet.
'And the world is responding, adopting Al faster than ever before. What all this progress means is that we're in a new phase of the Al platform shift. Where decades of research are now becoming reality for people, businesses and communities all over the world."