The Next Frontier: From Apps to Agents
For years, AI has optimized logistics and powered recommendation engines behind the scenes. Now, it's taking on a customer-facing role. The new goal is to create AI agents—digital assistants that do more than answer questions. They are being designed
to understand your needs from a simple conversational prompt and then take action, such as booking a multi-stop trip, ordering your weekly groceries, or planning a night out. Executives at Uber, DoorDash, and Airbnb see this as the next evolution of their platforms, transforming them from transactional tools into indispensable personal assistants. However, the vision is ambitious, and early data shows that widespread consumer adoption is not happening overnight.
The Reality Check of Early Adoption
Despite the industry buzz, executives have recently provided a dose of reality. During recent earnings calls, leaders from all three companies reported that the use of third-party AI agents to book services on their platforms is currently negligible. Uber's CFO, Balaji Krishnamurthy, revealed that during the busy Halloween week, AI agents booked a mere four rides and about 500 Uber Eats orders. Similarly, DoorDash CEO Tony Xu noted that third-party agents have only generated 'several hundred orders in total' since they've been active. Airbnb CEO Brian Chesky confirmed the trend, stating that bookings from external AI agents are 'very non-material'. This slow start suggests that while the long-term vision is powerful, current AI agents haven't yet earned consumer trust or proven more convenient than simply using the apps themselves.
Uber’s Vision for a Personal Concierge
Uber CEO Dara Khosrowshahi envisions an AI agent that acts as a 'personal concierge' for your life. The goal is to create a seamless experience where a user can conversationally ask the Uber app to plan an evening, which might involve booking a ride to a restaurant, reserving a table, and arranging the ride home. Khosrowshahi has noted that for simple, quick tasks—like the 50 seconds it takes to book a ride—users are unlikely to turn to an agent. However, for more complex planning, AI could stitch together Uber's various services. The company is also developing practical AI tools, like an Uber Eats 'Cart Assistance' feature that can turn a photo of a recipe into a grocery list, which has resulted in shopping carts twice the average size. While open to collaborating with third-party agents, Khosrowshahi is cautious, drawing on his past experience at Expedia to avoid becoming overly dependent on another company as a 'gatekeeper'.
DoorDash and the AI-Powered Commerce Engine
DoorDash CEO Tony Xu sees AI as the key to building a hyper-personalized 'local commerce engine'. Like Uber, DoorDash has found that third-party agent usage is currently very low, which Xu attributes to the fact that ordering a meal is already a simple task for most users. He believes AI agents will prove more valuable for complex, high-consideration purchases rather than simple food delivery. Internally, DoorDash has aggressively pursued AI, launching features like 'Ask DoorDash', which allows users to build grocery lists with natural language, and AI-powered voice ordering systems for restaurant partners. Xu's strategy is to embrace AI's potential as a future channel while acknowledging its current limitations, stating, 'You never want to fight gravity. But at the same time, you don't want to be delusional'.
Airbnb’s Ambitious AI Travel Agent
Of the three, Airbnb has perhaps the most ambitious and publicly discussed vision. CEO Brian Chesky plans to roll out an in-house AI travel agent in the first half of 2027 that helps users plan their entire trip. Chesky argues that travel planning is an aspirational and visual process that general-purpose chatbots are ill-suited for. Citing data that 80% of travelers enjoy the planning process, he says Airbnb's agent will be a 'custom-designed, AI-native experience' that assembles maps, photos, and itineraries rather than just responding with text. While external agent bookings are minimal, Airbnb is already seeing significant success with AI in other areas. Its AI-powered customer service assistant now resolves approximately 40% of all support issues without human intervention, leading to a 17% year-over-year decline in support costs per booking.
















