
Tech • AI • Robotics
Tesla, Waymo and Zoox have accelerated the shift from experimental self-driving cars to commercial robotaxi services, intensifying a race led by the United States and China while raising pressure on more cautious markets such as France.
Tesla has begun deploying its Cybercab service in Austin, presenting a vehicle designed from the ground up for autonomy, fully electric, lighter and intended to be cheaper to build. Waymo, owned by Alphabet, has unveiled a new generation of purpose-built vehicles rather than retrofitted cars. Zoox, backed by Amazon, has started its first paid service in Las Vegas with a steering-wheel-free shuttle layout featuring face-to-face seating.
The latest vehicles mark a break from earlier autonomous fleets based on modified production models such as Jaguar SUVs fitted with visible sensors. New platforms use flat floors, low step-in height and integrated doors to optimize passenger comfort and fleet economics. The result is a product aimed directly at mass urban transport rather than demonstration runs.
A central driver of the new momentum is lower hardware complexity and cheaper sensing systems. Waymo has reduced the cost of key equipment, including LiDAR, while improving integration. That shift matters because autonomy becomes disruptive only when a driverless ride approaches, then undercuts, the cost of a human-driven Uber or taxi.
Early adopters in San Francisco, Miami and Shenzhen describe robotaxis less as a novelty than as a superior transport product: personalized climate and music, privacy, and no driver-related friction. This comfort advantage is emerging alongside 24-hour availability, making autonomous fleets attractive not only on price but also on consistency.
Supporters of deployment point to large reductions in collisions. Figures cited for Waymo indicate about 80% fewer injury crashes and 95% fewer serious or fatal injury crashes than comparable human-driven travel. Those numbers, if sustained at scale, turn autonomy into a public-safety debate as much as a technology or industrial policy issue.
Commercial autonomy is already operating in parts of China and the United States, with Shenzhen frequently cited as an example of rapid deployment. In China, some autonomous rides have been offered at roughly €3 for 10 minutes, helped by scale and policy support. The market there extends beyond road vehicles, with autonomous aerial taxi trials also advancing.
The gap between technology companies and legacy manufacturers is becoming more visible. Systems from established brands such as BMW, Toyota, Ford, Cadillac, Mercedes-Benz or Renault are still largely driver-assistance tools, often limited to lane-following on highways. By contrast, Tesla and Waymo are developing full-stack autonomous platforms more akin to robotics and AI products than classic automobiles.
Even when software leadership sits in California, much of the hardware ecosystem depends on China and, for chips, also on Taiwan. Vehicles, sensors and electronics are often sourced or assembled through Chinese manufacturing networks. That makes robotaxis part of a wider geopolitical contest over who controls the digital and physical infrastructure of AI.
France and other European markets remain more restrictive, slowing access to advanced autonomous features. The concern is not only consumer delay but broader economic drag: slower gains in road safety, insurance savings, logistics efficiency and urban mobility. A late attempt to reserve the sector for incumbent automakers could deepen that lag if those groups cannot catch up technologically.
Autonomous taxis directly threaten drivers, but the discussion quickly extends to a wider pattern of AI-driven automation. The core political dilemma is whether governments prioritize preserving current jobs or accelerating technologies that improve safety and productivity. No clear consensus has emerged on how to manage retraining, income support or the speed of disruption if deployment scales quickly.
Robotaxis have entered a new commercial phase in which vehicle design, lower costs and safety metrics are beginning to align. The next divide may be less between companies than between countries willing to deploy the technology and those that hesitate.
Explain this