Autonomous vehicles have been perpetually two years away for about a decade. In 2026, they've arrived — but with so many asterisks that it's worth being precise about what "arrived" actually means. This week Motional's self-driving Ioniq 5 joined the Uber app in Las Vegas, available for hailing to and from casino resorts on the Strip, a shopping center, and Downtown. There is a safety monitor in the car. The service area is a handful of pre-mapped zones in one city. True driverless operation — no one in the vehicle at all — remains a stated 2026 goal rather than a current reality. This is progress, unambiguously. It is not the transportation revolution the press releases have been promising.
Motional's Las Vegas launch is particularly notable given the company's history. Two years ago it underwent a major operational reset, contracting sharply after burning through cash on a more ambitious deployment timeline. The return to Uber in a limited Vegas service represents a more disciplined approach: proven geofenced routes, safety drivers still aboard, a narrower commercial footprint. The Hyundai-backed company says true driverless service will follow before the end of 2026. The industry has heard that kind of timeline before, but Motional's reset arguably makes this version more credible than earlier iterations.
Uber's posture toward autonomous vehicles has become one of the more interesting strategic pivots in the industry. Rather than building its own AV technology — a path it abandoned when it sold its self-driving division — Uber is positioning itself as the distribution and data layer for the robotaxi ecosystem. Its new AV Labs division collects driving data for AV partners. Its Autonomous Solutions unit is pitching itself as a platform any robotaxi or delivery robot operator can plug into. The company has deals with Motional, Volkswagen (Los Angeles, 2026 launch), and Chinese operator Momenta (Europe). Uber's bet is that it doesn't need to own the technology; it just needs to own the customer relationship and the trip data.
China is moving faster and at larger scale. Pony AI and Toyota have begun commercial production of a driverless vehicle model, with Toyota's backing giving Pony the manufacturing muscle that most pure-play AV companies lack. Chinese robotaxi operators have been running without safety drivers in approved zones for longer than their US counterparts, partly because Chinese regulatory frameworks for AV deployment have moved more quickly than those in most US states. BloombergNEF counts highly automated vehicles operating in 103 cities globally — a number that would have seemed implausible five years ago, even if most of those deployments are modest in scope.
The honest assessment is that the robotaxi industry has found a viable initial form: small fleets, dense geofenced urban zones, premium positioning, and gradual safety-driver removal as data accumulates. That's meaningfully different from the mass-market autonomous commute that was the original vision. Whether it scales from that form into something economically transformative — or stays a niche service in a handful of cities — depends on whether the unit economics improve enough to justify expanding beyond the most controlled environments. That remains genuinely uncertain.