The robo taxi market is arriving in waves rather than all at once, and the Financial Times followed two of its central players to explain why that is. Waymo, owned by Google, has roughly 3,000 vehicles operating across more than ten US cities and has set its sights on London and Tokyo as its next expansion markets. Amazon-owned Zoox is pursuing a different path: its vehicle was designed from scratch solely for driverless urban use, can travel in either direction, and seats passengers facing each other in an enclosed cabin. Both companies are trying to capture a market that Uber's chief executive has described as a trillion-dollar opportunity. They are doing it with substantially different approaches to the hardware problem, and neither has yet demonstrated that the economics close at scale.
The pricing gap matters. In San Francisco, robo taxi fares are running approximately a third higher than an equivalent Uber ride with a human driver. That premium is a real constraint on adoption, and both Waymo and Zoox are aware that scaling volume is the only path to bringing costs down. The safety case, by contrast, is looking increasingly solid. Waymo's safety report, published in late 2025 and covering more than 127 million driverless miles, recorded no fatal crashes. The company's own analysis estimates a 90% reduction in incidents leading to serious injury compared to a hypothetical comparable fleet of human drivers. NHTSA estimates approximately one fatality per 100 million miles with a person behind the wheel. The caveats are real: the data still covers a limited set of operating environments, and a single high-profile incident could reshape public and regulatory perception quickly.
Zoox co-founder and CTO Jesse Levenson argues the economics of a purpose-built vehicle make more sense for full-day robo taxi operations than converting an existing car. A standard EV retrofitted for autonomy needs mid-day charging stops, which removes it from revenue service. Zoox's larger onboard battery is designed to eliminate depot runs during operating hours. The vehicle's airbag system also reflects its purpose: rather than deploying from the front, the bags envelop each passenger from the sides. Regulatory progress is the other variable. Bryant Walker Smith, an expert on autonomous vehicle law and policy, describes the current US approach as largely reactive: problem-specific and still building toward frameworks capable of handling the technology at city-wide scale. The UK has taken a more structured approach, with government-permitted Waymo trials in London confirmed for this year.
Bottom line: The technology is working well enough that the robo taxi market is real. Whether the economics and regulatory path align fast enough to justify the capital already committed is the question that determines which players are still standing in five years.