Almost no one uses AI agents to book their Uber, company says

During the week of Halloween, Uber handled around 320 million trips and orders across Uber and Uber Eats, but AI agents placed only four ride requests and 500 Uber Eats orders. CEO Dara Khosrowshahi described agent-driven transactions as having “very limited traction” so far, while calling AI agents a potential net positive for demand and noting that AI companies are in discussions with Uber. The figures suggest current adoption is small, with limited evidence yet that agents save customers time.

Analysis

The near-term read-through is less “AI agents are failing” than “there is little demonstrated urgency to replace a fast, habitual app flow.” That limits the case for immediate order-volume or customer-acquisition disruption at UBER and DASH, but also weakens the near-term monetization case for agent platforms. META’s reported app adoption is not evidence of transaction conversion; usage, completed commerce, repeat rate, and economics are separate milestones.

The longer-term risk is distribution control. If assistants become the default discovery and checkout layer, UBER and DASH could retain fulfillment while losing customer intent, interface engagement, and some ability to differentiate through their own apps. Conversely, a genuinely incremental agent-referred order could expand demand rather than cannibalize an existing app order. Which effect dominates depends on repeat behavior, fees, conversion, and whether platforms preserve control of checkout and customer data—none is established here.

Over days, this is weak evidence for a meaningful earnings revision; avoid treating it as a standalone short catalyst. Over 1–3 months, watch earnings commentary for agent-attributed orders and whether integrations are incremental. Over 6–18 months, the key structural signal is whether agents drive repeat commerce at attractive economics. Contrarian risk: current low usage may reflect early product maturity and minimal time saved on simple tasks, not a ceiling on agent adoption. The thesis changes if agent-referred orders scale materially, or if platforms disclose meaningful fees, conversion, or customer-acquisition effects.

Key Decisions for Investors

• No trade on this data alone: do not short UBER or DASH solely on weak early agent usage, and do not buy META as a commerce proxy based on app downloads. The reported activity is too early and the article supplies no evidence of material revenue or margin impact.
• For UBER and DASH, monitor quarterly disclosures for agent-sourced order share, repeat rate, incremental versus app-substituted orders, and any commercial terms with assistants. A rise in agent volume without evidence of incrementality or favorable economics would be a warning, not automatically a positive.
• Treat META and the AI-agent ecosystem, including Anthropic, Alphabet, and OpenAI, as an adoption watch rather than a transaction thesis: seek independently verifiable completed purchases and repeat usage before underwriting commerce monetization.
• Reassess the platform-disintermediation risk if assistants become a material source of completed orders and the platforms cede checkout or customer relationship control. Falsification of the near-term benign view would be company commentary indicating accelerating agent referrals alongside weaker app engagement, order frequency, or economics.