Hype Meets Reality: The Hard Numbers
Despite the excitement around a future where AI assistants book rides and order food, the reality of consumer adoption is strikingly modest. During its third-quarter earnings call, Uber revealed that in the typically busy Halloween week, third-party AI agents
booked only four rides and about 500 Uber Eats orders. This was out of a staggering 320 million combined transactions on the platform for that week. Similarly, DoorDash reported that since the capability went live, external AI agents have generated only a few hundred orders in total. For a company that can handle nearly a billion orders in a single quarter, this volume is negligible. Airbnb's CEO Brian Chesky has also described bookings from these external agents as 'very non-material'. The data makes it clear: consumers are not yet flocking to delegate their everyday tasks to these new autonomous tools.
The Efficiency Paradox
The reason for this slow adoption isn't a failure of technology, but a testament to the efficiency of existing apps. Uber CEO Dara Khosrowshahi noted that booking a ride on the Uber app already takes a median of just 50 seconds, while reordering a favorite meal on Uber Eats takes about 90 seconds. For these quick, high-frequency tasks, introducing an external AI agent doesn't solve a major problem; in some cases, it might even add a layer of complexity. DoorDash CEO Tony Xu echoed this sentiment, suggesting that consumers overwhelmingly stick to the native app for simple requests. He argued that AI agents are more likely to prove valuable for 'long consideration purchases' that require research and planning, rather than ordering a quick meal from a known restaurant. In short, it's hard to improve upon an experience that is already incredibly fast and simple.
Where AI Is Actually Winning
While third-party agents are getting a lukewarm reception, the real AI revolution is happening inside these company's own platforms. Here, AI is not being used to replace the user, but to enhance their experience and drive key business metrics. For example, Uber Eats has a feature called Cart Assistance, which uses AI to populate a grocery basket from a handwritten list or a recipe screenshot. Khosrowshahi said that carts built with this tool are roughly twice the average size of a normal basket. Likewise, DoorDash's internal tool, 'Ask DoorDash,' helps users discover new options with conversational search. The company reports that orders placed through this feature have an average basket value nearly 50% higher than conventional orders. This is the real story: AI is succeeding not as an external delegator, but as an internal co-pilot that helps users discover more and, ultimately, spend more.
Airbnb’s Vision for a Smarter Future
Airbnb has perhaps the most ambitious and integrated vision for AI. CEO Brian Chesky draws a sharp distinction between a simple 'chatbot' and a true 'agent' that understands you and your needs. While consumer-facing booking agents are still nascent, AI is already transforming the company from the inside out. Chesky credits AI with increasing the number of features and improvements the company ships by nearly 80% compared to the previous year. On the customer service front, its AI assistant now resolves nearly 45% of issues without any human intervention, dramatically lowering support costs. For Airbnb, AI is a foundational layer that accelerates productivity and improves efficiency across the entire business. Chesky has called it the 'best thing to ever happen to Airbnb,' framing it as a complete operational shift, not just a new feature.
The Strategic Long Game
So, what is the main point from these executives? It's a lesson in strategic patience. These platforms are not fighting the AI agent trend, but they are not beholden to its hype cycle either. They understand that for now, the real return on investment comes from using AI to perfect their own ecosystems. By making their internal tools smarter, they increase order sizes and improve operational efficiency. They view the eventual rise of powerful third-party agents not as a threat, but as another potential customer acquisition channel—one they are positioning themselves to control. As Khosrowshahi stated, all the major AI companies are talking to Uber to negotiate terms. This isn't a race to adopt external agents; it's a calculated strategy to own the fundamental 'rails of commerce' for when agent-led shopping truly arrives.
















