You might be feeling déjà vu.
A decade ago, companies like Google and Apple made promises that their respective voice assistants could complete tasks on your behalf. Apple said Siri could request an Uber—try it today, and Siri just opens the Uber app. Google claimed you’d be able to “Order my usual” at Starbucks with Google Assistant, but the experience was clunky, and Google eventually removed the feature.
But now, in the age of large language models and AI that understands natural language, we’re seeing the same promises back on the table. At Samsung’s Galaxy Unpacked event, Google and Samsung showed off how you’ll be able to use the Gemini voice assistant to complete tasks with a select number of third-party apps: book an Uber, or order food through Uber Eats, DoorDash, or Grubhub. (Expect more app support later this year when Android 17 arrives.)
The feature—still in early preview—will only be available in the US and South Korea to start, beginning with the Galaxy S26 smartphones when they launch on March 11, and later as a software update for the Google Pixel 10 series. Here’s how it works.
Step by Step
Ask Gemini to “Get me an Uber to the airport,” and you’ll see it open the Uber app in a virtual window. Gemini will continue to run this action in the background, but you’re able to monitor what it’s doing by tapping on the live notification. You’ll be able to see the entire process, though the idea is that you’ll just let it run in the background while you focus on something more important.
If it needs more information, say if you’re in the New York tristate area and it’s unsure which of the three major airports you want to go to, it’ll come back to you and ask those questions. Once it’s done, you’ll get a notification, and you’ll be taken to the Uber app’s booking stage. Gemini won’t make the booking decision for you, so you’ll have to choose whether you want an UberX or UberXL, confirm the fare, and tap book.
“I refer to some of the tasks that you might want to have automated as sort of digital laundry—things that you know you need to do, but are not necessarily excited about finishing,” Sameer Samat, president of the Android Ecosystem at Google, tells WIRED.
Samat showed me more complex examples of Task Automation with Grubhub. (At present, you can only run one automation at a time.) He opened up a group text thread, where a few friends talked about ordering pizza from Pizza Hut before a board game night. People mentioned specific pizzas in the thread. He called up Gemini, and asked it to, “Figure out the order.” With that, Gemini knew to get the context from the screen and organized everyones orders neatly.
Samat then asked Gemini to “order this on Grubhub for home delivery,” which then launched the task automation process. Gemini opened up a virtual window of Grubhub and got to work. A few minutes later, Gemini came back with all of the items in the cart. All Samat had to do was verify that everything looked good and hit order.
Instagram content
There are instances where Gemini may run into issues, but it will always try to figure out a workaround before coming back to you. After all, what’s the point of an assistant that keeps asking you questions when you just want it to do something on your behalf? He said once, Gemini came back to him asking what size and crust type he wanted. Another time, the restaurant limited the number of large pizzas he could order during a busy service, so Gemini came back and asked if two medium pizzas were OK instead.
Samat reiterates that Gemini hasn’t just memorized a “map” of the app—something we saw with early versions of AI agents, like Rabbit’s R1. Gemini uses its reasoning capabilities to make a plan, looks at the screen as you would, and navigates through it. Even if the app undergoes visual changes from today to tomorrow, Gemini can still figure out what to do. “It can be robust to things like that,” Samat says.
In another example, Samat showed me a Google Keep note with an RSVP list of people attending an upcoming barbecue, which also noted who was vegan. He called on Gemini to figure out how many hot dogs and buns were needed—Gemini did the calculation, and then Samat asked if Gemini could add the necessary items into his Safeway cart on DoorDash, and a few minutes later, it was all sitting in his cart.
AI Vision
Gemini’s task automation here is just one of three ways it can complete a task. If there’s a Model Context Protocol (MCP) integration—the open source universal language that lets LLMs talk to third-party apps—then Gemini can run the task in the backend. (In this instance, you wouldn’t see the whole process play out; you’d just see the final checkout step appear after you make the request.) There are also “App Functions” developers can build that allows Gemini to interface with it in a structured way. But if neither exist, then that’s when Gemini can open the app itself and navigate through the buttons, text boxes, and menus to complete tasks.
“This is the first time we’re doing this on Android with applications, and so getting this right is really important,” Samat says. “I think it’s an exciting step forward in technology. We sort of view this as the beginning of a new era of mobile intelligence, and Android is where we think you see the future first.”
Privacy concerns abound when it comes to granting Gemini access to your apps. Samat says that’s why Google hasn’t included any overly sensitive apps in this first batch for task automation. He says this data is not used for advertising, and that users can delete the data that Gemini sees. “We do think it’s really important that people have trust in the system, and that comes from having control and transparency of what it’s doing.”
While the smartphone screen is still required at the moment to complete the task, Samat envisions a future where you can start these tasks through other devices—say, a pair of smart glasses, an AI pendant, or even a car. (There will be several new Android XR-powered smart glasses launching this year.) He says the company is looking at other ways to make the final authentication capable on these other devices. After all, you wouldn’t want someone to steal your smart glasses and ask Gemini to order 55 burgers, 55 fries, 55 tacos—well, you get the idea.
The post Gemini Can Now Book You an Uber or Order a DoorDash Meal on Your Phone. Here’s How It Works appeared first on Wired.
