Gemini AI Gains Agentic Capabilities: Now Books Uber and Orders DoorDash

Google has taken a pivotal leap in artificial intelligence with the rollout of agentic capabilities to its Gemini AI assistant, allowing it to autonomously perform real-world tasks such as hailing Uber rides and compiling DoorDash orders. Initially available on the upcoming Pixel 10 and Samsung Galaxy S26 smartphones, this new functionality represents the first time a major AI assistant has been granted the ability to execute actions on behalf of users without requiring step-by-step manual input.

According to Wired, the feature, termed "task automation," activates when users issue a natural language prompt such as, "Book me an Uber to the airport by 7 a.m." or "Order my usual Thai curry from DoorDash." Gemini then interprets the request, accesses relevant apps, selects options based on user preferences and past behavior, confirms details, and completes the transaction—all without the user needing to open another app or manually confirm each step. This shift from conversational AI to autonomous agent marks a paradigm change in how users interact with digital assistants.

The implementation is being rolled out gradually, with initial availability limited to users in GDPR-compliant regions and select U.S. states, due to privacy and regulatory considerations. Google has emphasized that user consent is required at every stage: users must explicitly enable task automation in settings, and each action requires a final confirmation before execution. This layered consent model aims to balance convenience with user control, addressing concerns around AI overreach and data security.

Behind the scenes, Gemini leverages a combination of advanced natural language understanding, contextual memory, and secure API integrations with third-party services like Uber and DoorDash. The AI doesn’t just recognize intent—it maps it to actionable workflows, dynamically adapting to real-time variables such as traffic conditions, restaurant availability, and delivery fees. For example, if the user’s preferred restaurant is closed, Gemini may suggest alternatives based on past orders, dietary preferences, and delivery time estimates.

Samsung and Google have collaborated closely to integrate these capabilities into the Galaxy S26’s One UI, ensuring seamless interaction between Gemini and native Android services. On Pixel 10 devices, the feature is deeply embedded within Google’s ecosystem, pulling data from Google Calendar, Maps, and Payments to enhance personalization. Both manufacturers are also implementing on-device processing for sensitive data, reducing reliance on cloud servers and improving response times.

Industry analysts see this as a strategic move by Google to counter competitors like Apple’s Siri and Amazon’s Alexa, which have lagged in true task automation. "This isn’t just a feature update—it’s a redefinition of the assistant role," said Dr. Lena Torres, AI researcher at Stanford’s Human-Centered AI Lab. "Gemini is transitioning from answering questions to doing things. That’s the difference between a tool and a true agent."

However, the rollout has not been without scrutiny. Privacy advocates have raised concerns about the potential for unintended consequences, such as accidental orders or location tracking during task execution. Google has responded by providing users with an activity log that details every automated action taken by Gemini, along with the ability to delete logs or disable the feature entirely.

As task automation expands to other services—including flight bookings, calendar scheduling, and grocery shopping—Google aims to make Gemini the central nervous system of the user’s digital life. The next phase may include integration with smart home devices and car systems, further blurring the line between virtual assistant and personal agent. For now, with Uber and DoorDash as the first use cases, Google is proving that AI doesn’t just need to understand us—it needs to act on our behalf, intelligently and safely.