Uber: AI agents booked just 4 rides in its busiest week

Uber disclosed on its third-quarter 2026 earnings call, held on Wednesday, November 4, 2026, that only four rides were ordered through an AI agent during its busiest week of October, the week of Halloween, according to TechCrunch. An AI agent, in this context, is an assistant built by another company, a chatbot or personal assistant, that books a ride or a meal on the user's behalf without the user ever opening the Uber app.

Chief financial officer Balaji Krishnamurthy told analysts that Uber handled around 320 million trips and orders across Uber and Uber Eats during that week, according to TechCrunch, and that AI agents placed only about 500 Uber Eats orders. Crypto Briefing reports the same figures. On November 10, 2026, the Mexican outlet Expansión framed the same disclosure as a gap between the expectations of the tech industry and the habits of consumers.

How small is four rides against 320 million trips?

Adding the four rides and the roughly 500 food orders gives about 504 transactions placed by agents, or around 0.00016% of the 320 million trips and orders. That percentage is our own calculation from approximate figures, not a number published by Uber. Crypto Briefing computes the food channel alone and finds about one agent-placed order for every 640,000 trips and deliveries.

The sources are not perfectly aligned on the period covered. TechCrunch ties the four rides to the week of Halloween, while Crypto Briefing at one point speaks of the quarter. This article follows the TechCrunch account of the call, and the exact measurement window should be considered unconfirmed until Uber publishes a detailed breakdown.

It is very, very early and very limited traction that we are seeing so far.

Why are users not asking agents for a ride?

Khosrowshahi pointed out that booking a ride takes only 50 seconds on average in Uber's own app, and that re-ordering a favorite meal takes 90 seconds, according to TechCrunch. That suggests an AI agent does not necessarily save anyone time.

Cognativ, a software consultancy, lists further frictions as its own analysis rather than as statements from Uber: trust in the pickup point, price and time window, which pulls users back into the app; payment and identity permissions, which raise questions of liability; and habit, since people who open Uber daily have little reason to route the same task through another product.

Uber is not alone. DoorDash said on its own earnings call the week before that it had received only a few hundred orders from AI agents so far, because most customers still order through the DoorDash app, according to TechCrunch.

Is the hype around AI agents still justified?

TechCrunch notes that the figures arrive amid strong enthusiasm for agents that do more than answer questions and handle tasks for users. It cites the startup Instinct, founded last year, which raised its latest round at a valuation of 10 billion dollars. It also cites Meta's Muse, which became the number one app in Apple's US App Store in October and passed 10 million users according to Meta's own reporting, with many users shopping and transacting with businesses. OpenAI, Anthropic, Google and SpaceXAI (Grok) also offer their own agents, according to TechCrunch.

Is Uber still betting on agents?

Yes. Khosrowshahi said that agents are another opportunity to reach the Uber platform and that he expects them to be a net positive for demand across the ecosystem, according to TechCrunch. He noted that there are about eight major AI assistants on the market, expects more than ten in time, and added that every single AI system company is talking to Uber.

Cognativ adds, citing the call, that Uber is negotiating terms with assistant makers and wants to avoid depending on any single one. It also reports, as an early signal, that baskets built with Uber's own cart assistant are roughly twice the size of the average basket. This suggests the nearer-term value may sit in AI that a company controls inside its own product rather than in third-party agents acting from outside.

The core business remains strong. Crypto Briefing reports gross bookings of 59.8 billion dollars in the third quarter of 2026, up more than 20% year over year and the fifth consecutive quarter above that mark, and non-GAAP earnings per share of 0.88 dollars, up 35%. On autonomous vehicles, the same outlet reports a target of around 700 fully driverless vehicles on the network by the end of 2026 and more than 2,000 by the end of 2027. Its conclusion is that AI doing the driving has concrete targets, while AI doing the ordering has four rides.

One caveat on DiDi. The headline of the Expansión article says that the Chinese platform DiDi also acknowledges low adoption of AI assistants. The part of the article we could read does not detail DiDi's figures, so none are reported here.

What does this change for a company in Mexico?

None of the sources gives agent-booking figures specific to Mexico: the numbers above are Uber's global ones. They are still a useful reality check for Mexican companies in delivery, retail, mobility or services that are deciding how much to invest in being reachable by AI agents. A practical plan, drawn from the Cognativ analysis and Uber's own stance, has four parts:

• Put AI to work where you control the experience, such as customer support, quotes or scheduling, rather than waiting for outside assistants to bring you customers.
• Keep your systems ready for outside agents by documenting clean interfaces for search, pricing, booking and order status, with authentication and limits on automated traffic.
• Track agent traffic as a separate channel, so you can see when it starts to matter.
• Avoid locking yourself into a single assistant ecosystem, since the list of assistants is still growing and commercial terms are still being negotiated.

For a deeper look at how to deploy such tools inside a business, see our page on AI agents for business. The lesson of the Uber figures is one of sequencing, not skepticism: customers have not yet moved to agents even for simple, frequent tasks, so the near-term return is more likely to come from AI inside your own channels.