According to TechCrunch, Uber and Lyft have both announced plans to test Baidu’s Apollo Go robotaxi service in London starting in 2026. Lyft CEO David Risher stated on X that testing will begin once local regulators give the green light, with plans to eventually scale to “hundreds” of Baidu’s electric RT6 SUVs. Uber said it expects to start its own testing in the first half of 2026 as part of a deal with Baidu announced this past July. This move puts both U.S. ride-hailing firms alongside autonomous vehicle leaders Waymo and local UK startup Wayve, which will also have vehicles operating in the city next year.
The new ride-hailing playbook
Here’s the thing: Uber and Lyft aren’t really trying to build the robots anymore. That dream, with its massive capital burn, seems to have faded. Instead, their strategy is now about being the dominant platform, the app on your phone that hails whatever vehicle is available—human-driven or autonomous. They’re becoming aggregators of self-driving technology. This London deal with Baidu is just the latest example. Lyft has partnerships with Motional and Waymo. Uber works with Waymo and now Baidu. It’s a capital-light way to stay in the game without betting the company on a single, unproven tech stack.
Why this is a huge deal for Baidu
This isn’t just another test for Baidu. It’s a major geopolitical and technological beachhead. Getting its Apollo system onto the streets of a major Western capital like London is a massive credibility boost. For years, the narrative has been that China’s tech is walled off and doesn’t compete globally on cutting-edge software. A successful deployment in London, with all its complex weather, traffic, and regulations, would shatter that assumption. It positions Baidu not just as a Chinese player, but as a direct global competitor to Waymo and Cruise. That’s a big deal.
London is becoming the ultimate proving ground
So why London? It’s basically turning into the ultimate stress test for autonomous vehicles. Narrow, winding streets, relentless rain, aggressive cyclists, and famously complex traffic circles. If your system can work there, you can probably work anywhere. Having Waymo, Wayve, Uber-Baidu, and Lyft-Baidu all operating in the same city next year creates an incredible real-world experiment. We’ll get to see which approach works best, and the competition will accelerate development faster than any closed-course testing ever could. It’s a high-stakes race with a very public finish line.
What this means for the rest of us
Look, don’t expect to hail a driverless car in London next year and have it be cheap or ubiquitous. This is testing, likely with safety drivers, in limited zones. The real commercial launch is probably years away. But the direction is crystal clear. The ride-hailing model we know is on borrowed time. The endgame is a mixed fleet, and the companies that win will be the ones that best manage the transition from a human driver network to an automated one. For now, the biggest winner might be the city of London itself, which is positioning itself as the global hub for AV validation. Everyone wants to prove themselves there.
