Self-Driving Ubers Arrive in London — But Only 15 Cars Are on the Road

Londoners can now hail a self-driving taxi through the Uber app for the first time. The service, a partnership with UK-based AI firm Wayve, launched Thursday, but the chances of actually riding in one are slim. Both the BBC and The Guardian report that only 15 autonomous vehicles are part of the initial fleet, a tiny fraction of the more than 100,000 private hire vehicles operating in the capital.

This isn't the fully driverless future some might imagine. It’s a cautious first step. For Uber, it marks a return to the self-driving space in a new, lower-risk capacity after selling its own autonomous technology group years ago. Instead of building the cars, Uber is now positioning its app as the platform for other companies' technology. For Wayve, it's a chance to gather crucial real-world data and introduce its system to the public.

A Robotaxi With a Human Chaperone

Every self-driving Uber ride will come with a person in the driver's seat. All sources confirm these vehicles will have a safety driver ready to intervene if the car's AI system gets confused or makes a mistake. The BBC, whose reporter Zoe Kleinman took a ride in one of the vehicles, described an experience where the car capably navigated London traffic, but the human presence was a constant reminder that this technology is still in a supervised phase.

This approach stands in contrast to more aggressive rollouts in some US cities, where companies like Waymo operate fully driverless services without a human at the wheel. The London launch is a tightly controlled trial. It’s designed to test Wayve's AI in one of the world's most complicated driving environments while keeping the public and regulators comfortable. The goal isn't to take over the city overnight, but to prove the system can handle the unpredictability of London's streets, cyclists, and pedestrians.

More of a Trial Than a Takeover

With only 15 cars available, this launch is statistically insignificant for London's transport network. A user ordering an Uber has a very low probability of being matched with a Wayve vehicle. This confirms the project's true purpose: data collection. Wayve's AI learns directly from video data, so every mile driven in London—with its narrow lanes, complex roundabouts, and constant construction—is invaluable training material.

The partnership model is the key strategic shift for Uber. After spending and ultimately losing billions on its own in-house self-driving unit, Uber is now outsourcing the immense cost and risk of developing autonomous technology. By integrating with Wayve, Uber gets to offer a futuristic service and test the user experience of self-driving integration without the massive R&D overhead. It's a playbook that lets Uber remain a player in the autonomy race by becoming the indispensable platform that connects riders with whichever company's technology proves successful.