First, What Exactly Is an AI Agent?
Think of an AI agent as a digital employee that can do more than just answer a question. Unlike a simple chatbot that follows a script, an AI agent can understand a goal, make a plan, use different tools, and take action to complete a task on its own.
For example, instead of just telling you how to request a refund, an AI agent could check your order history, verify the store's policy, process the refund, and send you a confirmation. They are designed to handle multi-step processes that previously required a human, such as booking appointments, qualifying sales leads, or even managing complex logistics. This ability to reason and act is why major tech companies see them as the next frontier in customer interaction.
The 500 Uber Eats Orders: A Reality Check
The headline's mention of '500 Uber Eats orders' points to a specific and telling statistic. During Halloween week, a peak period with around 320 million combined trips and deliveries, external AI agents were used to place only about 500 orders on Uber Eats and book a mere four Uber rides. DoorDash reported similarly modest numbers, with CEO Tony Xu stating that third-party agents had generated 'maybe several hundred orders' in total since their launch. These figures, disclosed in recent earnings calls, show that despite the hype around new AI assistants from companies like Meta and OpenAI, consumer adoption for tasks like ordering food is still extremely low. Executives suggest a simple reason: it's already incredibly fast to use the apps directly, with Uber noting the median time to order a meal is just 90 seconds. For now, using a separate AI agent to do the job just isn't saving much time.
Uber and DoorDash: AI for Efficiency
While outside agents haven't made a dent, both Uber and DoorDash are heavily investing in their own internal AI. Their focus is on operational efficiency and enhancing the user experience. DoorDash, for instance, uses AI agents from Replicant to automate the process of phoning in orders to restaurants that aren't digitally integrated, handling tens of thousands of calls per day and even navigating noisy kitchens or voicemails. It also deploys AI-powered contact centers to manage the hundreds of thousands of daily support calls from customers, merchants, and its 'Dashers'. Similarly, both Uber Eats and DoorDash have introduced AI assistants that help users build shopping carts faster, sometimes by simply uploading a photo of a handwritten grocery list or recipe. This is where AI shines today: not by replacing the user, but by making complex or tedious parts of the service run more smoothly in the background.
Airbnb's Vision: The AI Travel Planner
Airbnb's approach is different, reflecting the nature of travel planning. While food delivery is about speed and convenience, booking a trip is often a more deliberate, visual, and even social process. CEO Brian Chesky has noted that users enjoy browsing photos and comparing options. As a result, Airbnb is developing a sophisticated AI agent, set to roll out in the first half of 2027, designed to act as an end-to-end trip planner. This agent will do more than just book a stay. The vision is for it to help with discovery, compare listings using dynamic interfaces with maps and photos, and manage the entire trip post-booking—assisting with changes, resolving payment issues, and providing support. Unlike the quick transactions of Uber and DoorDash, Airbnb sees its AI agent as a partner in a longer, more involved customer journey.
The Real Race: Owning the Customer Relationship
The low usage of third-party AI agents is actually good news for these platforms for now. It means that companies like Meta and OpenAI aren't yet succeeding in positioning themselves between the consumer and the service, which could have reduced Uber or Airbnb to mere fulfillment layers. However, all three companies understand the long-term potential and are building their own AI-native experiences to maintain their direct relationship with customers. They are also setting the terms for how external agents can access their platforms in the future. Uber's CEO, Dara Khosrowshahi, confirmed that 'every single AI system company' is in talks with them. This isn't just about exploring new technology; it's a strategic battle over who will guide consumer choices and own the valuable data that comes with it.
















