Starbucks has introduced a new beta experience within ChatGPT that allows customers to discover beverages conversationally and begin their orders directly from the chat interface, marking a shift toward AI-led commerce in the restaurant sector.
Launched on April 15, the integration enables users to interact with a Starbucks experience inside ChatGPT by describing moods, preferences, occasions or even uploading images to receive personalised drink suggestions. Instead of navigating a traditional menu, customers can prompt queries such as wanting a “light and refreshing morning drink” or “something indulgent for the evening,” with the system translating those cues into tailored recommendations.
Once a drink is suggested, users can customise it, adjusting milk options, sweetness levels or add-ons, before selecting a nearby store and initiating the order within ChatGPT. The transaction is then completed through Starbucks’ own app or website, ensuring that payments, loyalty integration and customer data remain within the company’s ecosystem.
Paul Riedel, Senior Vice President, Digital & Loyalty at Starbucks, said the experience reflects a change in how customers approach ordering. “Customers aren’t always starting with a menu, they’re starting with a feeling,” he noted, positioning the tool to meet users at the point of inspiration rather than decision.
The rollout builds on Starbucks’ broader push into artificial intelligence and personalisation, including its in-app recommendation engine, loyalty-driven offers and in-store tools such as “Green Dot Assist” designed to support baristas. The ChatGPT integration acts as a front-end discovery layer, encouraging experimentation with new or customised beverages while funnelling transactions back into Starbucks’ owned digital platforms.
The move also places Starbucks at the forefront of what is increasingly being described as “agentic commerce,” where AI platforms guide users from intent to transaction. Retailers and consumer brands globally are beginning to test similar models, but most, like Starbucks, continue to retain checkout within their own ecosystems rather than fully outsourcing the transaction layer.
Currently in beta, the feature is being used as a live testing ground to understand how conversational AI can influence ordering behaviour, increase menu exploration and drive higher engagement. Early feedback indicates strong interest in personalised recommendations, although broader consumer trust in AI-driven suggestions is still evolving.
