First, What Are AI Agents?
Before diving into the specifics, it’s crucial to understand what an 'AI agent' is. Think of it as a step beyond a simple chatbot. While a chatbot can answer questions, an autonomous AI agent is designed to understand a goal, create a plan, and take action
to achieve it with limited human supervision. Instead of just telling you about tacos, an AI agent could, in theory, know you like spicy food, check your favorite local spots, place an order for your usual, and have it delivered to your office, all from a single command. These systems are built to interact with different tools and platforms, make decisions, and complete multi-step tasks from start to finish.
The DoorDash and Uber Reality Check
The headline's mention of DoorDash and Uber reflects a new, albeit nascent, trend. Recently, both companies disclosed just how few orders are coming from third-party AI agents like those from Meta or OpenAI. During its busy Halloween week, Uber recorded only about 500 Uber Eats orders and a mere four rides placed by agents, out of 320 million total transactions. Similarly, DoorDash reported that since these agents launched, they've accounted for only 'several hundred orders' in total. These numbers are a tiny fraction of their overall business. The low adoption suggests that for tasks already made simple by well-designed apps—ordering a ride on Uber takes a median of 50 seconds—users don't yet see a need for an intermediary AI. However, both companies are still investing in their own AI tools, like text-to-order features, suggesting the strategy is to own the agent experience rather than ceding it to others.
Airbnb's Vision for an AI Travel Concierge
While DoorDash and Uber are seeing low external agent use, Airbnb is building its own ambitious version. The company announced plans to roll out an AI travel agent in the first half of 2027. Unlike a simple chatbot, Airbnb's agent is envisioned as a complete travel planner. CEO Brian Chesky has emphasized that travel planning is a visual and enjoyable process for many, something a text box can't replicate. Their agent will use a dynamic interface with maps, photos, and calendars to help users discover destinations and plan trips. Crucially, it will also handle post-booking logistics like making changes, processing refunds, and providing 24-hour support—tasks an outside agent can't easily manage. This strategy aims to create a specialized, indispensable tool rather than just another way to book.
Why Start Small? The Strategy Behind Cautious Rollouts
The low order counts from DoorDash and Uber aren't necessarily a sign of failure, but rather an indicator of a deliberate and cautious strategy. Deploying customer-facing AI on a small scale allows companies to manage risks, refine the technology, and handle errors before a wider rollout. Public concerns about AI agents 'going rogue' with credit card details or making ordering errors are valid, leading many users to be cautious. Furthermore, the true power of agents for businesses like Uber is currently being realized internally. Uber has successfully deployed AI agents to help its engineers with coding tasks like bug fixes and library upgrades, with reports indicating a significant portion of internal code is now AI-generated. This shows that while consumer-facing agents are in their infancy, agentic AI is already transforming complex internal workflows.
The Dawn of 'Agentic Commerce'
These early experiments are the first steps into what some are calling 'agentic commerce,' where AI takes on the role of a proactive economic actor on our behalf. The long-term vision is an ecosystem where your personal agent can find products, compare prices across different platforms, and make purchases. This presents both an opportunity and a threat to platforms like Uber and Airbnb. If they allow third-party agents to become the primary interface for customers, they risk becoming mere fulfillment layers, losing control over the customer relationship and data. By developing their own powerful, integrated agents, these companies are betting that specialized, in-house AI will provide a better experience that keeps users within their own ecosystems, setting the terms for how agentic commerce will evolve.
















