You might not notice the difference right away if you drive down South Congress Avenue in Austin on any given evening. Traffic flows. The lights shift. However, something quieter is taking place between the rideshare Priuses and the pickup trucks: a white car passes through an intersection with no one clearly behind the wheel. It doesn’t rubberneck, hesitate, or honk. All it does is drive. And depending on who you ask, that either feels like a step forward or like the beginning of something that hasn’t been fully considered yet.
For an industry that spent twenty years claiming it was nearly ready, Austin has emerged as an unlikely testing ground. It is home to Waymo, the autonomous car division that Alphabet has invested billions of dollars in over the past fifteen years. After months of preparation and Elon Musk’s typically ambitious timeline, Tesla launched its own robotaxi service in the city.
Key facts — Global robotaxi industry (2025–2026)
| Major players | Waymo (Alphabet), Tesla, Baidu Apollo Go, WeRide, Pony AI |
| Waymo weekly rides (Q2 2025) | ~250,000 per week |
| Waymo Q2 2025 California trips | 2.2 million (5× year-over-year) |
| Baidu Apollo Go weekly rides | ~250,000 per week (as of Nov 2025) |
| Tesla robotaxi status (Austin) | Launched 2025; human safety drivers still onboard |
| Waymo vehicle cost (est.) | Hundreds of thousands of dollars per unit |
| Chinese expansion markets | Dubai, Abu Dhabi, Singapore, Germany, UK, broader Europe |
| Waymo planned expansions | Washington DC, New York City, London |
| Notable incident (early 2026) | Dozens of Baidu robotaxis stalled mid-road in China, causing crashes |
| Reference / source | Waymo Research & Safety Reports |
However, its cars still have human safety drivers, which is either a sensible precaution or a subtle acknowledgement that complete autonomy is more difficult than the press releases implied. Austin was a conscious choice for both businesses. It is expansive, politically accommodating, conducive to technological experimentation, and has a reasonably predictable road layout. Put another way, it’s a controlled experiment dressed like a real city.
The fact that the competition is no longer limited to Americans is what makes this moment truly fascinating, if a little unsettling. With about 250,000 rides per week, Baidu’s Apollo Go, which has been quietly gaining rides in China for years, now matches Waymo in terms of weekly trip volume. It took Waymo more than ten years to reach that figure.
With goals that go far beyond Wuhan and Beijing, Baidu arrived sooner and at a fraction of the cost of the car. In Dubai and Abu Dhabi, Apollo Go cars are currently in service. WeRide has arrived in Singapore and the United Arab Emirates. The CFO of Pony AI told the Wall Street Journal that their hardware costs are “much, much lower” than Waymo’s. If this is accurate, it will alter the industry’s calculations in ways that American investors haven’t fully considered.
Observing all of this, it seems like the American robotaxi story and the Chinese robotaxi story have been running concurrently for so long that it will be shocking when they eventually collide in the same markets. WeRide, Baidu, and Pony AI have all announced plans for European operations, and Waymo is getting ready for its own London expansion, making the UK and Europe the first true overlap zone. In 2026, Uber-deployed Momenta vehicles are expected to begin operating in Germany. The question of whose safety record and regulatory relationships will hold up under scrutiny is genuinely open because so many autonomous vehicles will soon share the same small, wet streets.
Using transparency as a competitive tool has been Waymo’s approach. When Baidu’s figures were questioned, the company made this public and provided a wealth of safety data to US transportation authorities. Baidu asserts that in millions of miles of driving, its cars have never been involved in a serious collision. Perhaps this is the case, but it’s hard to evaluate without publicly available data. Baidu may have an impeccable record. It’s also possible that when auditing is minimal, maintaining immaculate records is simpler. That question was not resolved by the recent incident in which dozens of Apollo Go vehicles stalled simultaneously on Chinese roads and highways, some of which were allegedly involved in collisions.
It’s difficult to ignore the fact that the topic of robotaxis has changed from “will this ever work” to “who wins.” Even though the finish line is still hazy, that is a significant shift. In a single quarter last year, Waymo completed 2.2 million trips in California, five times more than during the same time in 2024. It is not a modest growth rate.
Nevertheless, the business continues to lose money, rely on Alphabet’s tolerance, and work toward a profitability that is declining at a rate that is only marginally quicker than the fleet’s growth. In the meantime, Tesla is marketing robotaxis as a fleet product—vehicles that owners can put on the network when they’re not using them—which is either a clever distribution strategy or a covert means of shifting the liability burden to customers.
Regardless of what transpires in Austin, London, or Dubai in the future, one thing is becoming more and more evident: the autonomous vehicle is no longer just an idea. It’s a business, a product, and now a geopolitical competition. For the most part, the technology functions. One intersection at a time, the more difficult issues—pricing, regulation, public trust, and which company’s definition of safety is truly safer—are still being worked out.
