Mobileye Global announced on June 16, 2026, that it will launch its own commercial robotaxi service in the United States in 2027. The company has spent years selling self-driving technology to other companies. Now it is entering the market as an operator — running its own fleet of robotaxis. That is a structural shift that changes how Mobileye competes.

The initial launch will cover approximately 100 vehicles in a single major U.S. city, according to Reuters. Mobileye's stated target is to scale that fleet to roughly 17,000 vehicles over the following five years. The service will run on Mobileye Drive, the company's full-stack driverless system — meaning all the software, sensors, and compute hardware packaged together — designed to power robotaxis without requiring a traditional car manufacturer as a middleman.

CEO Prof. Amnon Shashua signaled this direction at CES 2026, saying that robotaxi systems had matured enough to deploy commercially and that cheaper sensors would make the business economics work. The 2027 U.S. launch is the company following through on that message.

The Problem with Selling and Competing at the Same Time

Here is the core tension. Mobileye currently sells its self-driving technology to Volkswagen and MOIA, a mobility operator backed by VW, according to TechCrunch. By launching its own robotaxi service, Mobileye steps into direct competition with the same companies it supplies. You are both their technology vendor and their competitor.

This pattern appears elsewhere in tech history. Platform makers — companies that build technology others run on — have sometimes moved downstream into markets their customers occupy. Sometimes it works out. Sometimes it damages those customer relationships and pushes them toward alternative suppliers. Mobileye's parent company is Intel, which has navigated this exact dynamic across its own history. Whether Mobileye Drive is good enough that operator-customers accept the arrangement, or whether it accelerates their search for other suppliers, is unclear at this stage.

A Parallel Launch in Japan

Alongside the U.S. announcement, Mobileye and Japanese mobility operator WILLER have agreed to launch an autonomous robotaxi service in Japan. Full details on timing and terms were not disclosed, but the parallel track suggests Mobileye is pursuing robotaxi operations across multiple countries rather than testing a single U.S. city.

Japan's regulatory environment for autonomous vehicles is still evolving, with some early deployments already running in specific regions. A foreign technology company entering through a local operator partner like WILLER — which already runs intercity and regional transport services — fits how Japan typically prefers foreign companies to operate: under local accountability and partnership.

What the Numbers Actually Tell Us

Mobileye's plan to grow from 100 vehicles in year one to 17,000 over five years warrants a realistic assessment. One hundred vehicles is not a small pilot. It is large enough to generate real data on rare edge cases, maintenance requirements, and whether the business actually makes money per vehicle. But it is small enough to control liability and regulatory exposure in a single city.

The 17,000 target is a planning goal, not a binding commitment. Waymo, which has operated robotaxis longer than anyone else in the United States, spent years in Phoenix, San Francisco, and Los Angeles collecting operational data before expanding. Mobileye's growth plan assumes favorable regulatory approval and steady drops in the cost of sensors — particularly lidar, a key component for seeing the road — that may or may not occur on schedule.

Shashua's CES comments about sensor costs matter directly here. Robotaxi economics are not primarily limited by software right now. They are limited by hardware — the cost of lidar, cameras, and the computing unit in each vehicle. If Mobileye's internal roadmap on sensor costs holds, the 17,000-vehicle target becomes more achievable. If it slips, the business model becomes harder.

What the 2027 launch does establish is that Mobileye will hold operational and financial risk on its own books — not simply license its technology and collect per-vehicle fees from others. That commitment carries weight. Over the next two to three years, it will show whether Mobileye Drive can actually run a commercial robotaxi operation at scale without a large, well-funded parent company absorbing the losses and surprises that inevitably arise.