Mobileye is an Israeli company that most people have never heard of, even though its technology is inside the car they drive. For 25 years it made the computer chips and cameras that help cars detect pedestrians, stay in lanes, and brake automatically. Its systems are fitted in over 230 million vehicles from dozens of car brands. That is its business: sell the technology, let others do the driving.
On June 16, Mobileye said it is changing that. It plans to launch its own driverless taxi service in a major US city in 2027, starting with around 100 vehicles and scaling toward roughly 17,000 over five years if the first phase goes well. It will run the whole thing itself: the cars, the booking app, the fleet management, and the remote monitoring staff who watch the vehicles from a control room.
The logic makes sense on paper. A company that builds the brain of a self-driving car arguably knows better than anyone how to run one in commercial service. Controlling the whole operation also means Mobileye captures more of the revenue per ride instead of just selling components to someone else who captures it. Several other players, including Chinese automaker XPeng, are following the same thinking.
The problem is timing. Waymo, which is owned by Alphabet (Google's parent company), has been running paid driverless rides since 2020 and now completes around 500,000 of them every week across 10 US cities. It is targeting one million per week by the end of 2026. That is not a company Mobileye can easily catch. Waymo has years of real-world operating data, established city licenses, and a growing fleet that already includes over 3,000 vehicles.
Mobileye's own supplier history adds another layer of doubt. A robotaxi startup called Verne, backed by Croatian carmaker Rimac, originally planned to use Mobileye's self-driving platform for its service in Zagreb. Before the launch, Verne switched to a Chinese competitor called Pony.ai instead. That kind of slip is not a fatal sign, but it is the sort of thing investors and partners notice.
Mobileye's stock has fallen around 9% this year before today's announcement. The news sent shares up about 6%, which tells you investors think the move is at least the right direction, even if the destination is uncertain.
Running a fleet of driverless taxis is also a very different business from selling chips. Chips are high-margin: you design them once, manufacture them, sell them, and collect the payment. Running a taxi fleet means buying or leasing cars, paying for depots, cleaning, insurance, software updates, remote operators, and regulatory fees in every city you enter. Waymo is still working out how to make that economics stack profitable, which is why it recently launched a cheap monthly subscription plan to attract more frequent riders.
For businesses that use ride-hailing for staff travel, airport transfers, or last-mile logistics, none of this changes anything immediately. Driverless taxis in most cities are still either unavailable or limited. What is shifting is the pace. Two years ago this was a technology experiment. Today Waymo alone is moving half a million people per week without a driver. The Mobileye announcement is further evidence that multiple serious companies believe the business model works and are committing capital to prove it.
The city Mobileye will launch in matters a great deal and has not been announced. Regulatory approval processes vary hugely by city and state in the US. Mobileye says it will share more details at an investor event before the end of 2026. That event will tell us far more about whether this plan is credible than the announcement did.