You unlock your Pixel 10 Pro, tell Gemini you want pad thai, and watch it open DoorDash, tap through the menu, add your usual order, apply a promo code, and confirm—without you touching the screen. This is the moment that changes everything, even if right now it's slow, clunky, and limited to a handful of apps.
That's the user experience Google is shipping today on the Pixel 10 Pro and Galaxy S26 Ultra. Task automation puts Gemini in control of your installed apps, letting it complete real-world actions on your behalf. You set the goal; Gemini handles the execution. The Verge's hands-on testing confirms it works—you can watch the AI navigate Uber's interface, select a ride, and complete a booking while you observe. No demos. No keynote theater. This is actual functionality available to real users.
The scope is intentionally narrow. The Verge reviewer found the feature currently limited to food delivery and rideshare services—Doordash, Uber Eats, Uber. Google wants controlled conditions before expanding. The slowness is real: watching Gemini methodically tap through screens takes longer than doing it yourself. Occasionally it misinterprets a screen and needs correction. These aren't minor annoyances; they're the difference between a novelty and a replacement for your current workflow.
But here's what makes this significant: this is the first genuinely autonomous AI assistant on a phone. Previous assistants answered questions or controlled smart home devices. Gemini can operate third-party apps directly. It reads interfaces and takes actions. That's architecturally different from anything shipped before.
The competitive pressure is immense. Apple has been developing similar automation for iOS 19, with sources indicating a fall 2026 rollout. Samsung has established partnerships to bring Galaxy AI features to its devices. Google's head start on Pixel and Galaxy matters—if developers build integrations for Gemini's automation framework, Android becomes the platform where this capability arrived first.
The hardware requirement matters too. Task automation demands the processing power of the Pixel 10 Pro and Galaxy S26 Ultra, which start at $1,099 and $1,299 respectively. Gemini Advanced subscription adds $20 monthly. This is a premium experience for early adopters, not a mass-market feature yet.
What changes for you today: you can hand off specific tasks to Gemini and let it complete them while you do something else. The payoff isn't convenience in the short term—it's speed for some tasks, yes, but the real value is the foundation Google is building. Once the automation infrastructure exists, expanding to more apps becomes an engineering problem, not a research problem.
The Verge reviewer called it "impressive as hell" despite the limitations. After years of AI features that impressed in presentations but disappointed in practice, watching your phone actually complete a food order is a different kind of moment. Google has shipped something real. Whether it becomes indispensable depends on how fast the app integrations multiply and how quickly the clunkiness disappears.