The initial operating area covers selected locations in Jumeirah and Umm Suqeim, two coastal Dubai districts. The rollout does not cover the entire city. Uber and Baidu have said the operating area can expand as the service gains operational experience and receives further regulatory approvals.
That limited footprint is typical of an early robotaxi deployment: vehicles operate within approved zones rather than offering unrestricted service across every road in the emirate.
Riders in the launch area have three relevant ways to be matched with an autonomous vehicle:
The Uber option is therefore an additional booking channel, not a replacement for Apollo Go’s own app. The supplied launch reports do not specify whether autonomous rides have a different price from comparable trips with a human driver.
The vehicles are Baidu’s Apollo Go fully driverless robotaxis. They are designed to operate without a safety driver behind the wheel during the approved service. Dubai-based New Horizon Luxury Transport operates the local fleet.
This division of responsibilities is central to the business model: Baidu provides the autonomous-vehicle technology and Apollo Go platform, Uber provides customer access and ride matching, and a local operator handles fleet operations in Dubai.
The Uber launch followed a series of local development milestones:
The RTA’s role is more than a formal endorsement. Its approvals determine where autonomous vehicles can test and operate, creating the route from controlled validation to commercial passenger service.
Uber and Baidu announced their multiyear partnership in July 2025. The agreement aims to place thousands of Apollo Go autonomous vehicles on Uber’s platform across multiple international markets outside the United States and mainland China. Dubai is the first market where that agreement has reached a rider-facing deployment.
For Baidu, Uber offers an established customer-acquisition and booking network in markets where Apollo Go does not yet have a large local consumer platform. For Uber, partnering with multiple autonomous-vehicle companies can expand driverless supply without requiring Uber to develop every autonomous-driving system itself.
Baidu’s existing operational scale is part of the rationale for taking Apollo Go abroad. In the first quarter of 2026, Apollo Go reported 3.2 million fully driverless operational rides, with weekly rides peaking above 350,000 in March; total rides were up more than 120% year over year.
Apollo Go is not the only autonomous service available through Uber in Dubai. Uber and WeRide launched passenger robotaxi rides in parts of Jumeirah and Umm Suqeim in December 2025, followed by fully driverless fare-charging operations in March 2026.
Uber’s Dubai strategy consequently looks less like a commitment to one autonomous-vehicle technology and more like a marketplace approach. The platform can connect riders with vehicles from different technology providers, while each partner supplies its own autonomous-driving stack and operational model. Uber has also announced robotaxi partnerships with companies including Pony.ai for European deployments.
For Baidu, the competitive challenge is different: Apollo Go must demonstrate that its experience in China can translate into safe, reliable operations under different road conditions, regulations, fleet arrangements, and customer expectations.
The public announcements emphasize ride volumes, international partnerships, and fleet expansion. They do not disclose Dubai-specific revenue, per-ride costs, cash burn, pricing differences, or the revenue-sharing terms between Uber and Baidu.
That makes it premature to describe the Dubai service as profitable. The launch demonstrates commercial availability and partnership execution, but it does not by itself establish attractive unit economics. The key evidence to watch will be fleet utilization, operating-zone expansion, pricing, repeat usage, and whether the partners publish results from the Dubai deployment.
Apollo Go plans to expand its Dubai fleet to more than 1,000 vehicles, subject to regulatory approvals and operational progress.
The likely near-term direction is a phased increase in fleet size, operating zones, and Uber rider matches rather than an immediate citywide service. Expansion will depend on RTA permissions, vehicle readiness, safety performance, and the capacity of the local operator.
Baidu is also testing the international model beyond Dubai. Apollo Go began open-road testing in London with Uber and Lyft in July 2026, while public passenger rides there remain subject to regulatory approval and are not expected to begin until 2027 according to the cited reporting.
The Dubai launch is therefore best understood as a live test of a scalable model: Baidu supplies driverless vehicles, Uber supplies demand, and local authorities determine how quickly the service can move from selected districts to a broader urban network.
The initial operating area covers selected locations in Jumeirah and Umm Suqeim, two coastal Dubai districts. The rollout does not cover the entire city. Uber and Baidu have said the operating area can expand as the service gains operational experience and receives further regulatory approvals.
That limited footprint is typical of an early robotaxi deployment: vehicles operate within approved zones rather than offering unrestricted service across every road in the emirate.
Riders in the launch area have three relevant ways to be matched with an autonomous vehicle:
The Uber option is therefore an additional booking channel, not a replacement for Apollo Go’s own app. The supplied launch reports do not specify whether autonomous rides have a different price from comparable trips with a human driver.
The vehicles are Baidu’s Apollo Go fully driverless robotaxis. They are designed to operate without a safety driver behind the wheel during the approved service. Dubai-based New Horizon Luxury Transport operates the local fleet.
This division of responsibilities is central to the business model: Baidu provides the autonomous-vehicle technology and Apollo Go platform, Uber provides customer access and ride matching, and a local operator handles fleet operations in Dubai.
The Uber launch followed a series of local development milestones:
The RTA’s role is more than a formal endorsement. Its approvals determine where autonomous vehicles can test and operate, creating the route from controlled validation to commercial passenger service.
Uber and Baidu announced their multiyear partnership in July 2025. The agreement aims to place thousands of Apollo Go autonomous vehicles on Uber’s platform across multiple international markets outside the United States and mainland China. Dubai is the first market where that agreement has reached a rider-facing deployment.
For Baidu, Uber offers an established customer-acquisition and booking network in markets where Apollo Go does not yet have a large local consumer platform. For Uber, partnering with multiple autonomous-vehicle companies can expand driverless supply without requiring Uber to develop every autonomous-driving system itself.
Baidu’s existing operational scale is part of the rationale for taking Apollo Go abroad. In the first quarter of 2026, Apollo Go reported 3.2 million fully driverless operational rides, with weekly rides peaking above 350,000 in March; total rides were up more than 120% year over year.
Apollo Go is not the only autonomous service available through Uber in Dubai. Uber and WeRide launched passenger robotaxi rides in parts of Jumeirah and Umm Suqeim in December 2025, followed by fully driverless fare-charging operations in March 2026.
Uber’s Dubai strategy consequently looks less like a commitment to one autonomous-vehicle technology and more like a marketplace approach. The platform can connect riders with vehicles from different technology providers, while each partner supplies its own autonomous-driving stack and operational model. Uber has also announced robotaxi partnerships with companies including Pony.ai for European deployments.
For Baidu, the competitive challenge is different: Apollo Go must demonstrate that its experience in China can translate into safe, reliable operations under different road conditions, regulations, fleet arrangements, and customer expectations.
The public announcements emphasize ride volumes, international partnerships, and fleet expansion. They do not disclose Dubai-specific revenue, per-ride costs, cash burn, pricing differences, or the revenue-sharing terms between Uber and Baidu.
That makes it premature to describe the Dubai service as profitable. The launch demonstrates commercial availability and partnership execution, but it does not by itself establish attractive unit economics. The key evidence to watch will be fleet utilization, operating-zone expansion, pricing, repeat usage, and whether the partners publish results from the Dubai deployment.
Apollo Go plans to expand its Dubai fleet to more than 1,000 vehicles, subject to regulatory approvals and operational progress.
The likely near-term direction is a phased increase in fleet size, operating zones, and Uber rider matches rather than an immediate citywide service. Expansion will depend on RTA permissions, vehicle readiness, safety performance, and the capacity of the local operator.
Baidu is also testing the international model beyond Dubai. Apollo Go began open-road testing in London with Uber and Lyft in July 2026, while public passenger rides there remain subject to regulatory approval and are not expected to begin until 2027 according to the cited reporting.
The Dubai launch is therefore best understood as a live test of a scalable model: Baidu supplies driverless vehicles, Uber supplies demand, and local authorities determine how quickly the service can move from selected districts to a broader urban network.