Copy article

The London Robotaxi Land Grab: When "Safer AI" Meets the Rush to Own the Road

ended 23. December 2025

Uber and Lyft will begin testing Baidu's Apollo Go robotaxis in London from the first half of 2026, pending regulatory approval. Lyft CEO David Risher confirmed testing will start with dozens of vehicles, with plans to scale to hundreds. Uber announced testing in the first half of 2026 as part of its partnership with Baidu announced in July

The UK government now aims to begin permitting robotaxis to operate in small-scale pilots starting in spring 2026. This follows the Automated Vehicles Act 2024, which aims to facilitate commercial pilots of services with paying passengers and no safety driver from spring 2026. 

The vehicles are Baidu's RT6 model, which costs 204,600 yuan to manufacture (approximately £22,000) , more than 50% less than the previous generation. The RT6 is equipped with 5 LiDAR units and 40 sensors across 7 categories, with computing power of 1200 TOPS. 

For Lyft, the 2026 pilot marks the first operational result of its £155 million acquisition of the FreeNow mobility platform, finalised in July 2025, which gave Lyft access to taxi licences in London. 

Waymo is also preparing to test in London, with plans for full service launch in 2026. British startup Wayve has partnered with Uber for UK trials.

Baidu's Apollo Go completed 1.1 million rides in Q4 2024, representing a 36% increase from the previous year. Baidu operates 250,000 weekly robotaxi trips across 22 cities worldwide. 

Industry analysts note that listed autonomous firms like Pony.ai and WeRide continue to report operational losses, £48 million and £34 million respectively for Q3 2025, despite surging revenues.  

The UK government launched a 12-week consultation on safety principles in June 2025, with regulations expected to come into effect in early 2026. 

The tech giants are selling "innovation," but for London's 120,000 drivers, it looks more like a high-speed eviction.

The Mayor of London has made a promise to work with the taxi industry and unions to minimise the impact of driverless vehicle technology on jobs and livelihoods. 

Mayor Sadiq Khan stated in a written response to a parliamentory question published 4 December 2025: “While we want London to be the global capital of innovation, enterprise and creativity, this must not be at the expense of jobs and livelihoods, and I am committed to working with industry and unions to ensure this technology is introduced in a way that works for all Londoners.”

DM News, a specialist publication for the private hire and black cab industry, reported in November 2025 that cab drivers were calling for protection as robotaxi operators prepare to launch in the capital.

We'd like your views:

  • What happens to existing professional drivers when low-cost robotaxis enter the market at scale?
  • How should the UK balance rapid deployment timelines with comprehensive safety validation for autonomous vehicles?
  • What transparency standards should apply to AI systems operating on public roads, particularly regarding training data and decision-making processes?
  • Should there be mandatory requirements for autonomous vehicles to be trained on local road conditions before deployment?
  • How can regulators ensure passenger safety when multiple competing robotaxi operators launch simultaneously?
  • What safeguards are needed when ride-hailing platforms operate vehicles they don't manufacture or directly control?
  • How should data residency requirements work for autonomous vehicles using AI trained in different jurisdictions?

4 responses from the Newspage community

Copy all

Star Quote
Copy

What safeguards are needed when ride-hailing platforms operate vehicles they don't manufacture or directly control? In San Francisco, a prankster summoned 50 Waymo robotaxis to a dead-end street simultaneously creating a 10-minute log jam of veicles in the road proving these systems are vulnerable to deliberate misuse, not just technical failures.

Baidu's vehicles use 5G remote control capabilities allowing human operators to take over in emergencies, but the UK's data residency requirements only specify that mapping and passenger data must be stored on UK servers, and nothing about where the remote control centre operates or who has access.

When your robotaxi fleet can be deployed by anyone with a phone and malicious intent, and the accountability chain runs through three countries before reaching the customer, you're not building transport infrastructure, you're building complexity without clarity on who's actually in control
Copy

If I’m climbing into a cab on my own at night, I’ll take the vetted human over the Terminator brain, thanks. Look, I love a bit of tech. But my personal safety isn’t a product launch. A London black cab driver has had checks, licensing, and real-world accountability. If something goes wrong, there’s a person with a name, a badge, and consequences. That matters when it’s late, you’re tired, and you just want to get home without turning it into a true crime podcast. A robotaxi might be great at not speeding or getting distracted. Fine. But women don’t only worry about crashes. We worry about the weird stop, the doors not unlocking, the route going off, and who helps when you hit panic mode. “Press here for support” is not the same as a human sorting it out fast. So yes, the robot might be statistically safer on paper. But until it proves proper emergency controls, rapid human backup, and clear UK accountability, I’m sticking with the checked-and-licensed cabbie.
Copy

Technology and automation almost always lead to job losses, certainly in the short term. Robo-taxis are therefore clearly bad news for existing taxi drivers. If the cost savings from removing human drivers are passed on to consumers through lower fares, the implications could be far broader — potentially leading to reduced car ownership.

We are already seeing some early signs of this. Many younger people are delaying learning to drive, influenced by cost, urban living and alternative transport options. For example, I see this with my own two boys, who are quite happy to use Ubers, even here in Leicester, and feel no real urgency to learn to drive.
Copy

Uber and Lyft aren't trying to fix London's congestion; they are automating it to pad their margins. Replacing experienced drivers, who navigate London’s complex ‘urban canyons’ by instinct, with cost-cutting robots that get confused by narrow alleys and tall buildings will only choke the city further.

That is why Mayor Khan’s intervention is so vital. He is absolutely right to stand with the unions and demand that innovation doesn't come at the cost of livelihoods. While the tech giants view drivers as 'legacy costs' to be deleted, the Mayor is correctly recognising them as the backbone of London's transport.

There is zero plan from these platforms to upskill cabbies; the plan is simply to replace them and pocket the difference when surge pricing hits. We must hold the Mayor to his promise: technology should serve London’s workforce, not drive them off the road for a quick profit.