Uber Earnings Data Reveals AI Agents Struggle to Gain Real-World Transactional Traction
Uber reported extremely low utilization of personal AI agents for ride-hailing and food delivery during its third-quarter earnings call. As reported by TechCrunch, the transportation giant processed only four AI-initiated rides during the busy Halloween week, despite industry-wide hype surrounding agentic technology and high-valuation startups entering the space.
Context from Industry Pulse
The disconnect between AI agent valuations and actual consumer usage highlights a significant hurdle for the agentic web. While Meta’s Muse app recently reached the top of the U.S. App Store with 10 million users and Instinct secured a $10 billion valuation, these milestones have not yet translated into high-volume commerce for service platforms. Uber and DoorDash both report that traditional apps remain faster or more intuitive for routine tasks. This data suggests that while the infrastructure for AI-driven commerce is being built by giants like SpaceXAI and Anthropic, the consumer habit of manual app interaction remains deeply entrenched due to the high efficiency of current mobile interfaces.
During a recent financial update, Uber executives shared data indicating that autonomous software assistants are not yet a mainstream method for accessing their services. According to TechCrunch, the company recorded a mere four rides booked via AI agents during the week of Halloween, which is typically one of their highest-volume periods. While the Uber Eats delivery segment saw slightly more activity with 500 AI-assisted orders, these figures represent a negligible fraction of the 320 million total trips and orders completed during that seven-day window.
This lack of adoption persists even as the technology sector sees a surge in investment for task-oriented AI. TechCrunch notes that startups like Instinct have reached multibillion-dollar valuations and established players like Meta, Google, and OpenAI have launched their own agentic tools. However, the convenience of existing mobile applications remains a significant barrier to change. Uber leadership observed that booking a ride manually takes less than a minute, while re-ordering food takes about ninety seconds, suggesting that current AI tools do not yet provide a meaningful time-saving advantage for consumers.
Uber is not the only platform seeing this trend. As TechCrunch reports, DoorDash recently disclosed similar results, citing only a few hundred orders placed through AI agents. Most customers in the gig economy continue to rely on traditional app interfaces for their daily needs. Despite the current low engagement, Uber management expressed optimism that these systems could eventually serve as a new channel for demand once the technology matures and gains wider acceptance.
The competitive landscape for these assistants is expanding rapidly. According to TechCrunch, there are currently eight major AI assistants available, a number expected to grow beyond ten in the near future. Uber confirmed it is in discussions with every major AI system developer, positioning itself to integrate with whatever platforms eventually gain consumer trust, even if the current transactional volume remains in its infancy.
The Industry Pulse Bottom Line
Uber's data proves that despite massive capital investment and developer interest, AI agents currently fail to offer a compelling enough efficiency gain to disrupt established mobile booking habits for most users.