Access all our platforms from a single account.
Robotaxi reality check: Tesla’s big swing meets a crowded road
Key takeaways
Tesla’s robotaxi story shifts from promises to pilots, but proof still matters more than posts.
Waymo and China’s leaders scale faster today, while Uber turns robotaxis into a marketplace game.
Traditional car makers lag in software and data, but some can still compete through partnerships and assisted driving.
The robotaxi dream returns every few months like a sequel nobody asked for, and yet here we are, watching the trailer again. This time, the catalyst is concrete: on 15 December 2025, Reuters reports Tesla is testing robotaxis without a “safety monitor” in the front passenger seat, after launching a limited, geo-fenced service in Austin in June. A geo-fence simply means the cars operate only inside a defined area, not “anywhere, anytime”. At the prior close, on 15 December 2025, Tesla closed at 475.31 USD, up 3.6%, on the day. The market reaction matters because autonomy is not just a feature upgrade. It is the potential engine of a very different business model. Competitors are already on the road: Alphabet’s Waymo runs paid robotaxi services, Baidu’s Apollo Go scales across Chinese cities, Uber positions itself as the booking layer, and General Motors has reset Cruise ambitions after setbacks. In robotaxis, the “winner” is less about the slickest demo and more about running safe fleets at scale, city by city.
Tesla’s bet: scale first, polish later
Tesla’s pitch is simple enough for a billboard: millions of cars, one software brain, and a future where your car earns money while you sleep. The latest testing update matters because it suggests Tesla is pushing beyond supervised trials. Reuters notes Tesla is testing with no occupants in the car and is preparing for a “Cybercab” launch next year.
Tesla’s advantage is also straightforward: fleet scale and data. Even if “Full Self-Driving” (FSD) is not fully autonomous today, Tesla collects huge volumes of real-world driving footage. That can help train systems to handle the boring parts of driving, plus the weird parts that make humans sigh and insurance companies sweat.
The hard part is that robotaxis are not a video game where you patch the bugs after launch. Regulators and city officials need confidence that the system fails safely, not creatively. In October 2025, Reuters reports the US National Highway Traffic Safety Administration opened a probe into nearly 2.9 million Tesla vehicles over alleged traffic violations and crashes linked to FSD use. That kind of scrutiny is normal in this space, but it slows timelines and raises the cost of proving safety.
The competition is already giving paid rides
If Tesla is the loudest story, Waymo is the quietest proof. Reuters reports Alphabet’s Waymo leads with more than 2,500 commercial robotaxis across major US cities as of November, and a media report cited around 450,000 paid rides per week.
China’s leaders are also scaling, and they are doing it with a different playbook: more city-by-city rollouts, more partnerships, and often a tighter link between local regulators and operators. Reuters reports that Baidu’s Apollo Go has a fleet of more than 1,000 fully driverless vehicles across 15 cities, has completed more than 11 million rides as of May 2025, and is now pushing international deployments via partners.
This is where Uber becomes an important character. Uber is not building the “driving brain” anymore. Instead, it is positioning itself as the demand layer. Reuters reports Uber and Baidu plan to deploy thousands of Apollo Go vehicles on Uber’s platform in international markets outside the US and mainland China, with initial rollouts expected in Asia and the Middle East.
In plain English: if robotaxis become real, the customer relationship and the app placement may matter almost as much as the sensors. Think less “car company” and more “distribution shelf”.
Why traditional car makers struggle, and the few ways they still matter
Most traditional car makers are excellent at building cars. Robotaxis require excellence at building a software system that drives, learns, and gets audited. That is a different muscle, and it is expensive to build late.
There is also an awkward business logic problem. A robotaxi aims to replace many privately owned cars with fewer, highly used cars. That can be great for the robotaxi operator, but it can be uncomfortable for companies whose profits still depend on selling more vehicles.
Some incumbents have already stepped back from standalone robotaxi ambitions. In December 2024, Reuters reports General Motors said it would end robotaxi development at Cruise, citing the time and resources needed to scale.
The exceptions usually follow one of two paths. First, premium and safety-led brands that focus on limited, well-defined autonomy features (for example, hands-off driving in specific conditions) rather than full robotaxis. Second, manufacturers that accept they will not own the whole stack and instead partner with specialist autonomy providers and compute suppliers. Nvidia sits in that “picks and shovels” role, supplying the computing hardware and software tools many developers rely on. The value may accrue to several layers, not just the car badge.
Risks that investors should actually watch
Robotaxis are a safety product first and a revenue product second. A single high-profile incident can pause deployment, trigger investigations, and change consumer trust. Regulation can also fragment the market into hundreds of local rulebooks, which makes scaling slower than a global smartphone app.
Economics is the second risk. The robotaxi future needs high utilisation, low downtime, and a cost per mile that beats a human driver, after maintenance, remote support, mapping, and insurance. If utilisation disappoints, the model starts to look more like a very smart, very expensive taxi.
Finally, competition risk is real. If multiple networks operate in the same city, pricing pressure rises, and profits can end up thinner than the “platform” narrative suggests.
Investor playbook
- If permits expand city-by-city, treat it as “real adoption”, not just better demos.
- If safety scrutiny rises, watch for slower rollouts and higher compliance costs across the sector.
- If partnerships multiply, follow who owns the customer, the fleet, and the economics in each deal.
- If robotaxis scale, reassess traditional car demand assumptions, especially in dense urban markets.
When the trailer becomes a film
Robotaxis always sell a simple idea: transport becomes cheaper, safer, and available on tap. The messy part is that driving is not one problem. It is a million small problems, in bad weather, with roadworks, cyclists, and human unpredictability.
Tesla’s latest testing shift moves the story from “someday” toward “show me”, which is progress. But Waymo and China’s leaders remind us that the winners may be the firms that treat autonomy as a slow, regulated rollout, not a single big reveal. For investors, the sensible stance is not blind belief or blanket scepticism. It is watching the proof points, city by city, mile by mile, until the trailer finally becomes a film.
This material is marketing content and should not be regarded as investment advice. Trading financial instruments carries risks and historic performance is not a guarantee of future results.
The instrument(s) referenced in this content may be issued by a partner, from whom Saxo receives promotional fees, payment or retrocessions. While Saxo may receive compensation from these partnerships, all content is created with the aim of providing clients with valuable information and options.