The Israeli autonomous driving supplier, best known for its early work with Tesla's Autopilot, says it will run a US robotaxi service by 2027, putting it in competition with the automakers and mobility platforms that already buy its technology.
Mobileye spent two decades building the perception software and driver-assistance systems that automakers bolt onto their own vehicles. Now the Israeli autonomous-driving supplier is preparing to compete with those same automakers by running its own US robotaxi service.
The company said this week that it will launch a vertically integrated robotaxi business in 2027 in an as-yet-unnamed American city, with an initial fleet of roughly 100 vehicles and a longer-stated ambition of scaling into the tens of thousands of robotaxis within five years. Mobileye will use its Moovit mobility platform to handle booking, ride coordination, and dispatch, building the operational layer alongside the autonomy stack it already sells.
That is a sharper break than it looks. Mobileye has spent years positioning itself as a Tier-1 supplier, the kind of vendor that sells SuperVision and Mobileye Drive systems to automakers and mobility platforms such as Volkswagen's MOIA, Lyft, Polestar, and Porsche, none of which want to compete with the company wiring up their cars. By deciding to operate its own service, Mobileye is choosing to compete directly with the customers it depends on.
Prof. Amnon Shashua, Mobileye's founder and CEO, framed the move as a way to combine the company's AV technology stack with operational ownership. "This initiative is going to accelerate adoption, gain direct operational experience, and create a financially and geographically scalable business," Shashua said in the company's announcement, as reported by Ars Technica.
The pitch is that running a fleet teaches Mobileye things that selling boxes to OEMs cannot. The risk is that OEMs and mobility platforms, some of them still Mobileye customers, read the same announcement as a sign that the supplier is now also their rival.
Mobileye has been here before, in a different costume. The company first rose to prominence as the supplier of the vision system behind Tesla's early Autopilot system. That relationship ended in 2016 after Mobileye publicly raised concerns about how Tesla was using the technology, a split that left Mobileye dependent on a broader set of automakers. Intel acquired Mobileye in 2017 and took it public again in 2022, according to the same Ars Technica report.
The 2027 service is announced, not operating. The target city is unnamed. The fleet size is company guidance, not deployed vehicles. And the company has not yet confirmed which automaker platform will supply the robotaxis, identified regulators it is working with, or laid out the operational design domain for the service.
The same playbook has been announced repeatedly across the AV industry. A Tier-1 supplier promises to vertically integrate, sketches a multi-thousand-vehicle roadmap, and pins the launch to "as soon as" dates that slip. Mobileye's prior 2025 plan to operate in Dallas through Lyft has not materialized as a commercial service. The new announcement extends that pattern with a 2027 target and a still-unnamed city.
What makes this announcement analytically different is the supplier-to-customer overlap. If Mobileye's robotaxi business does scale, the question is not just whether the 100-vehicle pilot grows into a multi-thousand-vehicle fleet. It is whether automakers and mobility platforms that buy Mobileye's technology today will keep buying it from a company that is also bidding for their riders.