Uber fulfilled about 320 million trips across rides and Uber Eats in Halloween week, and four of the rides came from third-party AI agents, chief financial officer Balaji Krishnamurthy said on Uber’s third-quarter call on 4 November, according to a published machine transcript.
He put Uber Eats orders from agents at around 500. By our arithmetic, about 504 agent transactions is roughly one in 635,000 of the week’s trips, or about one second of its average trip flow.
Business Pill · Agentic Commerce
An evergreen Business Engineer explainer on agentic commerce: what changes when the shopper is an AI agent. It explains the general idea only and does not discuss Uber.
The key insight:Uber counted four rides and around 500 Uber Eats orders from third-party AI agents in a Halloween week of about 320 million trips, its chief financial officer said on the third-quarter call; by our arithmetic that is about one in 635,000, roughly one second of the week’s average trip flow.
The figures came in the last answer of the call, after J.P. Morgan’s Doug Anmuth asked about AI agents. Chief executive Dara Khosrowshahi turned to Krishnamurthy, who said Uber had had “one of our largest weeks last week because it was Halloween”.
For Uber Eats, he said, the volumes coming from AI agents were around 500 orders in total. For rides, he said, “it was a grand total of four trips that we fulfilled from the agents”. He added: “So it’s very, very early and very limited traction that we’re seeing so far”.
Uber did not name the agents in that answer. A second account of the call, a summary on BigGo Finance, gives the same figures: 500 delivery orders and four rides in a recent week, against 320 million total trips.
A grid of 1,000 dots for about 320 million Uber and Uber Eats trips in Halloween week, each dot about 320,000 trips. The ~504 orders and rides from third-party AI agents that Uber’s CFO described on the Q3 call fill about 0.16% of one dot. The shares are our arithmetic.
What Was Posted on X
Uber recorded ~320 million rides and Uber Eats orders during Halloween week. How many rides came from AI agents Just FOUR. Yes, 4.
$UBER CFO also says AI agents generated just 500 Uber Eats orders.
Looks like the agentic commerce revolution still has a long way to go.
The Transcript’s post at 16:16Z on 4 November, which had about 2,000 views within about 20 minutes, summarised the CFO’s answer: about 320 million rides and Uber Eats orders in Halloween week, four rides and about 500 Eats orders from AI agents.
$UBER CEO says they will not surrender their direct customer relationships to AI intermediaries
“Listen, I used to work in the travel business where we were dependent on the Google of the world. I don’t want to go back there. Really the best-in-class marketplace technology that we’re building, operational execution, I think is going to have more and more people coming direct to us. We are building our own AI experiences.”
The Transcript’s post at 16:29Z quoted the chief executive’s travel-business remark. Its wording differs slightly from the machine transcript we quote, which has ‘the Googles of the world’.
Uber and UberEats processed 320MM trips over Halloween week.
Eric Seufert (@eric_seufert) posted at 15:51 UTC on 4 November 2026 that Uber and Uber Eats processed 320MM trips over Halloween week, that 500 orders were made through an agent and that four rides were booked through an agent. His conclusion, “Agentic commerce is a mirage”, is his opinion.
One Second of the Week
Our arithmetic: about 500 Eats orders plus four rides is about 504 agent transactions. Divided into about 320 million trips, that is roughly one in 635,000.
A week has 604,800 seconds, so 320 million trips is about 529 a second on average. On that average, the agent transactions add up to about one second of the week’s trips.
I Don’t Want to Go Back There
Before the figures, Khosrowshahi said Uber’s focus is to “invest for the best possible experience to become direct”. He said members account for more than 50% of bookings and that 75% of rides are a single tap.
He then said: “I used to work in a travel business where we’re dependent on the Googles of the world. I don’t want to go back there”. He also said that “every single AI system company is talking to us” and that Uber is “negotiating the terms of trade”.
He said that “there are eight different assistants” and that he expects more than 10. He called third-party agents “another opportunity to reach our platform” and said: “We think it’ll be a net positive in terms of demand into the ecosystem”.
What the Prepared Remarks Say
Uber’s prepared remarks give the median time to book a ride as about 50 seconds and to reorder food as about 90 seconds. They say Cart Assistant baskets are roughly twice the size of the average Grocery & Retail order, and that Uber launched a beta of Ask Eats, a voice assistant.
The remarks say: “We will continue to invest in AI experiences within our own apps while selectively working with third-party agents that can bring incremental demand”. They add that Uber’s operational capabilities “will remain valuable regardless of where consumers initiate transactions”.
Against DoorDash’s Count
On 29 October DoorDash chief executive Tony Xu said DoorDash was seeing “maybe several hundred orders in total from some of these third-party agents”, as we reported in Amazon, DoorDash CEOs Say Few Shoppers Buy Through AI Agents. By our arithmetic, 1,000 orders is about one in a million of DoorDash’s 993 million third-quarter orders.
As we read the call, Uber treats outside agents as a channel to negotiate with: its chief executive said every AI system company is talking to Uber and that it is negotiating the terms of trade.
As we read the two calls, DoorDash and Uber gave the same kind of answer six days apart: a count of agent orders in the hundreds or fewer, alongside an emphasis on their own apps.
Dara Khosrowshahi, Uber chief executive, Q3 call, 4 Nov 2026
“I used to work in a travel business where we’re dependent on the Googles of the world. I don’t want to go back there”
Three Implications
FOR AI AGENT MAKERS Khosrowshahi said there are eight different assistants and that he expects more than 10; Uber’s prepared remarks say it will work selectively with third-party agents that can bring incremental demand.
FOR UBER’S OWN APPS The prepared remarks put the median time to book a ride at about 50 seconds and say Cart Assistant baskets are roughly twice the size of the average Grocery & Retail order.
FOR DELIVERY RIVALS DoorDash counted several hundred agent orders in total on 29 October, about one in a million of its third-quarter orders by our arithmetic; Uber gave its figures alongside one week’s trips.
The framework says: “Pricing power leaves the model and lands on whoever controls the surface and its reusable workflows.”
As we read it, Uber’s chief executive is describing a contest over that surface: he wants riders and eaters to start in Uber’s own app, while Uber negotiates terms with the assistants that want to sit in front of it. The framework predates this call and does not discuss Uber.
Uber had not posted its official call transcript when we checked at about 16:40Z, so the quotes come from a published machine transcript, which contains speech-recognition errors; we quote only passages that read cleanly. Wording differs slightly between transcripts.
Krishnamurthy gave the agent figures in the same answer as the Halloween-week total, and one summary says they cover a recent week, but Uber did not state the period in writing. The figures are approximate, the agents are unnamed, and the DoorDash comparison is not like-for-like. We make no forecast about agent volumes.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Uber said about 320 million trips ran through its apps in Halloween week, with four rides and around 500 Eats orders from third-party AI agents, about one in 635,000 by our arithmetic. Its chief executive, recalling the travel business’s dependence on the Googles of the world, said: “I don’t want to go back there”.
Related coverage (4 Nov): the rest of Uber’s quarter, from its driverless-car plans to Mobility revenue, is in our Q3 earnings piece.
A note on sourcing. Call quotes come from a published machine transcript of Uber’s third-quarter call on 4 November 2026 (Uber’s official transcript was not yet posted), checked against a second account of the call; Uber’s prepared remarks are its official document. DoorDash figures are from our earlier coverage. The ~504 total, one in 635,000 and one-second comparisons are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
![This Week In Business AI: The $195B Month [Week #9-2026]](https://i0.wp.com/fourweekmba.com/wp-content/plugins/contextual-related-posts/default.png?resize=150%2C150&ssl=1 "This Week In Business AI: The $195B Month [Week #9-2026]")
