Wayve and Uber launch UK’s first autonomous ride-hailing service in London

Wayve and Uber launched the UK’s first autonomous ride-hailing service in London on September 3, 2026. The vehicles are Ford Mustang Mach-E electric vehicles equipped with Wayve’s autonomous driving system, the Wayve AI Driver, and sensors that detect their surroundings. When users request a ride through the Uber app and choose UberX, Uber Electric or Uber Comfort, an autonomous vehicle may be assigned. There is no additional charge. For now, the service operates under supervision, with a safety driver certified by Transport for London (TfL) in the vehicle.

According to Wayve, users can request an autonomous vehicle through the existing ride-booking process in the Uber app. This means consumers do not need to learn how to use a separate service, while presenting Uber with the operational challenge of integrating autonomous vehicles into its ride network. Wayve says it has trained its technology in London since 2018, underscoring the importance of accumulating urban driving data. A key question going forward is how much real-world operational data can be used for training under the requirement for an onboard safety driver.

The focus is not on driverless operation itself, but on integrating autonomous vehicles into ride-hailing across London. Rather than using a dedicated app or a limited service area, Wayve and Uber have inserted autonomous vehicles into the existing booking flows for UberX, Uber Electric and Uber Comfort. This can be seen as a move beyond showcasing test vehicles to treating autonomous driving as part of a ride-booking platform’s operations. However, the number of vehicles has not been disclosed, and service remains limited in scale as it operates under supervision with safety drivers on board. A related article in this newspaper, NHTSA grants Zoox its first commercial exemption for vehicles without steering wheels or brake pedals, examined how regulators handle purpose-built driverless vehicles under commercial exemptions. In contrast, Wayve and Uber are, at least for now, taking an approach that reconciles regulation and real-world operations by keeping TfL-certified safety drivers on board. Rather than confronting regulatory barriers head-on, the companies are building a track record through supervised operations. This illustrates the difference between two paths for autonomous driving: deploying driverless vehicles through regulatory exemptions, and integrating them into existing ride services under supervision. On the technology front, Wayve’s claim that it has trained in London since 2018 is significant. If its AV2.0 approach is to avoid reliance on HD maps and hand-coded rules, accumulated experience in complex urban environments will be a core asset. The launch is therefore more than a service announcement: it will test whether Wayve can create a cycle that collects London driving data through operations and uses it in subsequent training. The pace of fleet expansion, the conditions for assigning safety drivers and the range of driving conditions in which the vehicles will be allowed to operate autonomously remain unclear.

According to Wayve, Uber designed the in-vehicle experience, which includes an interactive screen supporting 64 languages. Passengers can use the screen to start a trip and check the planned route. Wayve is headquartered in London and says it has trained its AI in the city’s complex road environments since 2018. The company also says its “AV2.0” approach uses an AI Driver that learns from experience rather than relying on HD maps or hand-coded rules. The companies have not disclosed the number of vehicles in the service.

Wayve says its AV2.0 approach does not rely on HD maps or hand-coded rules. In Japan, too, accumulating urban driving and operational data is a prerequisite for autonomous driving, and integrating ride-hailing apps with vehicle control is likely to be a key issue.