Field report SE-40 / Q9 · sheet /strategy

Corporate StrategyFiled from the territory

AI Agents Ordered Just Four Uber Rides in the Company's Busiest Week

Uber logged just four agent-driven rides out of 320 million trips in a single week, and DoorDash has seen only several hundred AI orders — the hype has not yet reached the checkout.

Grid
SE-40 / Q9
Surveyed
November 17, 2026
Reading
3 min read
Correspondent
Elena Vasquez
Almost nobody uses AI agents to order food or rides on apps like Uber
Fig. SE-40 / Q9Almost nobody uses AI agents to order food or rides on apps like Uber — Icsilviu / pixabay

Field notes

  1. Only about 500 of Uber's 320 million trips in the last week of October 2026 came from AI agents — including just four ride-hailing trips.
  2. DoorDash CEO Tony Xu said the company has received 'several hundred orders in total' from AI agents, against billions delivered annually.
  3. Uber CFO Balaji Krishnamurthy called agent traction 'very, very early, and very limited'.
  4. DoorDash began testing a text-to-order feature in September 2026 that bypasses its own app.
  5. Shoppers cite ordering errors and credit card misuse fears as reasons to limit agent use.

Autonomous AI agents accounted for just four ride-hailing trips on Uber during the last week of October — out of 320 million trips the company completed in that period. The disclosure, made by Uber CFO Balaji Krishnamurthy on the company's third-quarter earnings call on Wednesday, is the sharpest quantification yet of how little traction AI agents have gained in mainstream commerce despite a year of industry predictions that they would transform online ordering.

How small are the numbers?

They are vanishingly small. Uber completed 320 million ride-hailing and delivery trips in the final week of October — a period Krishnamurthy called "one of our largest weeks," driven by a Halloween surge. Of those transactions, roughly 500 Uber Eats orders came through AI agents such as Meta's Muse and Instinct, alongside the four ride-hailing trips. Put in per-day terms, agents generated on the order of 70 food orders daily across Uber's entire platform — a rounding error against the millions of rides and deliveries Uber processes every 24 hours.

Krishnamurthy did not dress up the figures. "It is very, very early, and very limited traction that we have seen so far," he told analysts on the call.

DoorDash's numbers point in the same direction. CEO Tony Xu said a week earlier that his company had received "several hundred orders in total" from AI agents — a sliver of the billions of orders DoorDash delivers each year. Even at the top of that range, agent-driven demand would amount to well under one order per thousand across the platform's annual volume.

Why are consumers holding back?

The hesitation is practical, not ideological. Some AI users have already handed agents responsibility for purchases spanning plane tickets to groceries, and shoppers say the tools are genuinely useful for monitoring prices. But two categories of friction have kept adoption narrow, as Business Insider has previously reported:

  • Ordering errors. Agents have made mistakes in live transactions, prompting users to keep them on a short leash.
  • Payment risk. Fears that an agent could go rogue with credit card details have led cautious shoppers to limit what the tools are allowed to do.

The pattern that emerges from shopper accounts is a division of labor: agents are trusted to watch, compare and recommend, while humans still confirm and pay.

What are the platforms doing about it?

Both companies are building anyway. Limited usage has not stopped Uber or DoorDash from pressing ahead with AI tools inside their own apps. DoorDash said in September it was testing a text-to-order feature that lets users place orders without opening the company's app at all — a signal that the platform sees conversational ordering as a strategic interface even before external agents send meaningful volume its way.

The strategic logic is straightforward. If agent-mediated commerce eventually scales, the marketplaces that have already wired their systems to accept machine-placed orders stand to capture that flow; those that wait risk being disintermediated. For now, the experiments cost little relative to the option they preserve.

What does this mean for the agent economy?

The gap between narrative and usage is wide enough to measure. Industry forecasts have positioned AI agents as the next interface for e-commerce, yet the two largest on-demand platforms in the United States are collectively logging agent volumes in the hundreds against transaction volumes in the billions. Early adopters exist, and the technology works often enough to keep them — but the guardrails users demand, and the error rates they tolerate, have not yet converged at a point where agents can carry a meaningful share of daily commerce.

Four rides in the busiest week of the year is the number that frames the moment.

via feeds.businessinsider.com (Source)

Share this article:

More from Elena Vasquez

Elena Vasquez

Show full bio

Staff writer covering companies, deals and strategy in corporate strategy and chief-executive leadership at CEO Frontier.

486 articles

Filed under

Adjacent sheets

« Previous articleNext article »

End of sheetSE-40 / Q9 · November 17, 2026