AiGpu

Industry·

Uber Reports Minimal AI Agent Usage for Ride and Food Orders

Uber's Q3 earnings call shows only a handful of rides and a few hundred food orders were placed via AI agents, highlighting the early stage of consumer‑facing agent adoption.

Uber app interface showing ride booking screen with AI agent icon

Uber's third‑quarter earnings call revealed that AI agents accounted for only four ride requests and roughly 500 Uber Eats orders during the busiest week in October, the week of Halloween.

CEO Dara Khosrowshahi described the traction as "very, very early" and noted that the average booking time in the Uber app is about 50 seconds for rides and 90 seconds for food, limiting the time‑saving appeal of personal agents.

Despite the low numbers, Khosrowshahi remains optimistic, viewing agents as an additional demand channel and expecting the number of major AI assistants to grow beyond ten.

Why it matters for GPU and AI infrastructure

The modest transaction volume signals that consumer‑facing agent workloads are still tiny, meaning the immediate GPU demand from inference for personal agents remains limited. However, as platforms integrate agent APIs at scale, the backend inference load could rise sharply, creating future capacity planning considerations for GPU cloud providers.

Monitoring early adoption metrics helps infrastructure teams size clusters for the next wave of agent‑driven traffic.

  • aigpu
  • ai gpu
  • ai gpu cloud
  • aigpu dubai
  • uber
  • ai agents
  • infrastructure
  • industry

By AiGpu Editorial · Editorial rewrite based on public reporting (TechCrunch AI)

← All articles