The chip and software supplier is launching a 100 vehicle US robotaxi service in 2027, scaling to roughly 17,000 over five years, a vertical integration bet that puts it in direct competition with VW, Ford, and other customers it has spent years courting.
For most of its public life, Mobileye has been an arms dealer to the self-driving industry. It built the chips and the software that let other companies claim they had a credible autonomous vehicle program, then collected royalties and per-vehicle fees as those programs shipped. That business model is about to collide with the company's own next move: a US robotaxi service it plans to launch in 2027, starting with about 100 vehicles in an unnamed American city and scaling to roughly 17,000 over the following five years. The supplier is becoming the operator, and the operator will sit across the table from the same automakers that have been paying Mobileye's bills.
The shift, reported by TechCrunch's Kirsten Korosec on Monday, is more than a product expansion. Mobileye, the Intel-owned chip company that trades on Nasdaq under the ticker MBLY, has spent the last decade building a tier-one supplier business on top of its EyeQ family of perception processors and three overlapping software stacks: SuperVision, for advanced driver-assistance features on consumer cars; Chauffeur, the higher-end "eyes-off" highway product now reaching production; and Drive, the higher-level autonomous platform designed for robotaxi and B2B fleet buyers. Each of those products is sold to automakers, not run by Mobileye. The robotaxi service breaks that pattern, and it does so on American roads where two incumbents already operate.
Waymo, the Alphabet-owned autonomous vehicle unit, runs commercial robotaxi service in multiple US cities. Tesla has a robotaxi product live in Austin and is expanding. Mobileye's entry will be small by either comparison, but the strategic signal is bigger than the fleet size. Chief executive Amnon Shashua, in Mobileye's announcement carried by TechCrunch, said the "robotaxi revolution has only just begun" and argued that the industry "has become increasingly dependent on a small number of technology providers and business models." The second half of that sentence is the more revealing one. Mobileye is betting that, in a market where the software layer has consolidated into a few vendors, the operators who can capture per-ride margin will be the ones that own both the technology and the service. Going direct lets the company keep economics it would otherwise split with a fleet partner.
The tension is that several of Mobileye's biggest customers are also building their own mobility services. Volkswagen, through its MOIA subsidiary, runs a commercial ride-pooling operation in Hamburg and Hannover and has been one of Mobileye's anchor partners on the Chauffeur program. Ford, another publicly named Mobileye customer, ran its own Argo AI robotaxi effort before shutting that program down. Geely and Polestar round out the list of automakers Mobileye has named. The company is now telegraphing that it intends to compete for US robotaxi riders against any service those same automakers might want to launch. That is a real channel conflict, and the conflict is not hypothetical: the same Mobileye stack that VW uses for its consumer advanced driver-assistance work is a credible starting point for any VW-branded mobility service the automaker might want to operate in the United States.
The financial geometry makes the move more pointed. Mobileye is publicly traded, which means its investors have to be told about the strategy in a way that a private supplier does not. Its robotaxi service will compete for capital, engineering attention, and regulatory goodwill with the same automakers that supply a large share of its chip and software revenue. The company can frame the move as additive, another customer vertical in Mobileye's own words alongside its consumer and B2B programs, but additive businesses at suppliers-turned-operators have a long history of becoming awkward ones. Compaq selling PCs direct while also supplying components to Dell is the classic cautionary tale. Mobileye's leadership will have to answer, eventually, why an automaker should trust its autonomy roadmap to a vendor that is also its competitor.
There are also gaps in the announcement that will matter for how the story plays out. Mobileye has not named the launch city, the vehicle platform, or the operating partner for the 2027 service. The company has also not said whether the robotaxi fleet will run on Mobileye-owned vehicles, vehicles purchased from an automaker customer, or a mix. Each of those answers will reshape the conflict calculus. A fleet of Mobileye-branded Geely minivans is a very different competitive posture than a fleet of Volkswagen ID vehicles, and the choice will tell investors and customers how serious the company is about treating its own service as a real business rather than a reference deployment.
The broader question is what the move says about where autonomous mobility value is concentrating. If the bet works, Mobileye will have staked a claim to both ends of the autonomy stack, the perception-and-planning software layer it already owns and the per-ride service revenue it has not yet touched. If it does not, the company will still own the supplier business, but it will have burned capital and customer trust to find out. Either outcome reshapes how automakers think about the autonomy vendors they choose to depend on. The supplier model that built Mobileye is now visibly incomplete, and the company has decided to test the operator model on the most competitive roads in the world.