Dubai's Robotaxis Are Already on the Road. Here's What Happens Next
Dubai's Robotaxis Are Already on the Road. Here's What Happens Next
If you've been walking around Umm Suqeim or Jumeirah lately and spotted a car with no one gripping the wheel while it glides through a roundabout, you weren't imagining it. Since December 12, 2025, Dubai residents and visitors have been able to open the Uber app, tap an "Autonomous" option, and get matched with a WeRide robotaxi — making Dubai one of a small handful of cities outside China and the US where you can actually hail a self-driving car today.
It's a genuine milestone, but it's also easy to overstate. Here's what's actually happening, who's behind it, and how far the city still has to go before "driverless" means what it sounds like.
What's actually on the road right now
The service currently running through Uber isn't fully driverless. Each WeRide vehicle has what the company calls a "vehicle specialist" sitting in the driver's seat — not actively steering, but there to intervene if the autonomous system needs help. That's a deliberate, staged approach: let the software handle real Dubai traffic — aggressive lane changes, pedestrians darting between parked cars, roundabouts — with a human safety net, and only remove that net once the data justifies it.
RTA and WeRide have said the goal is to move to a fully driverless commercial service in early 2026, which would make this one of the shorter supervised-to-unsupervised windows of any robotaxi rollout anywhere. For context, Waymo spent years running supervised and limited unsupervised pilots in Phoenix before scaling to multiple US cities.
The geographic footprint is still narrow by design. Rides are limited to Umm Suqeim and Jumeirah — two coastal, well-mapped neighborhoods with wide roads, heavy tourist and resident foot traffic, and relatively predictable layouts compared with, say, Deira's older street grid. That's typical for how AV operators de-risk a launch: master a geofenced area before expanding it.
Who's actually building this
The partnership stack is worth untangling, because several names are doing different jobs:
- WeRide is the Guangzhou-founded autonomous driving company supplying the self-driving technology and vehicles. It's one of the most internationally active Chinese AV firms, already running services or pilots in cities across the Middle East, Europe, and Asia.
- Uber provides the booking layer — the same app used for ordinary rides now has an "Autonomous" toggle in supported Dubai zones, so riders don't need a separate app to try the service.
- Tawasul is the local fleet operator, handling the vehicles on the ground — maintenance, operations, and the practical business of running a fleet in Dubai's climate and regulatory environment.
- RTA (Roads and Transport Authority) is the regulator and strategic sponsor, issuing permits and setting the pace of expansion.
RTA hasn't put all its chips on one partner, either. Alongside WeRide, it has run trials involving Baidu's Apollo Go autonomous driving platform, with reporting suggesting roughly 50 Baidu vehicles and 10 WeRide vehicles involved in on-road trials feeding into the 2026 plans. RTA has also signed a memorandum of understanding with Pony.ai, another major Chinese autonomous driving company, signaling it wants multiple technology providers competing and complementing each other rather than one exclusive vendor controlling the city's self-driving future.
The number that matters more than the launch date
The headline isn't really "robotaxis exist in Dubai now" — it's the target behind them. RTA's Self-Driving Transport Strategy, first laid out back in 2017, aims to convert 25% of all journeys in Dubai into autonomous trips by 2030. Crucially, that figure isn't just about taxis: it spans buses, trains, marine transport, and private vehicles too, not only the robotaxi pilots making headlines.
That framing matters because it explains why RTA is moving so deliberately on partnerships and geography. A quarter of all trips in a city of roughly 3.5 million people is an enormous systems problem — it requires autonomous tech that works reliably across multiple transport modes, insurance and liability frameworks that regulators are comfortable with, and public trust that only builds gradually, ride by ride, in neighborhoods where nothing goes wrong.
What to actually watch for in 2026
If you want a signal that's more meaningful than another press release, watch for three things: whether WeRide's vehicle specialists actually come out of the driver's seat as promised, whether the geofenced zones expand beyond Umm Suqeim and Jumeirah into denser parts of the city, and whether Baidu or Pony.ai vehicles start showing up on consumer-facing apps the way WeRide's already have.
Dubai has a track record of announcing ambitious tech timelines and then genuinely hitting most of them, from the metro's driverless trains to its early embrace of blockchain for government paperwork. The robotaxi push looks like it's following the same playbook: pick a visible, contained pilot, get the human-in-the-loop version working flawlessly first, then widen the circle. Whether "early 2026" holds for the fully driverless version is the thing worth checking back on — not whether Dubai has robotaxis, because as of today, it already does.