Uber will once again operate its own robotaxis—this time with Rivian’s not-yet-built EVs

(SeaPRwire) –   Following a six-year hiatus, Uber is evidently striving to reintroduce its own robotaxis, utilizing agreement structures that appear intended to mitigate its exposure to risk.

This marks a major shift from several years back, when Uber divested its self-driving division, ATG, subsequent to a deadly 2018 accident and prolonged periods of substantial financial deficits. In the interim, Uber has pursued an alternative strategy—securing partnerships with almost every significant robotaxi entity in the industry, ranging from Waymo to WeRide. Tesla remains the sole exception that does not collaborate with the ride-hailing firm, although this was not due to a lack of effort from Uber.

Throughout these agreements, Uber has incorporated the autonomous vehicle fleets of other firms into its application; the AV companies retain ownership and operation of the automobiles. This dynamic is now shifting.

Initially, there was an agreement with Lucid Motors in 2025 to acquire and roll out as many as 20,000 vehicles featuring Nuro’s autonomous technology. On Thursday, Uber revealed a comparable arrangement with Rivian concerning its forthcoming R2 platform. The firm intends to buy 10,000 fully autonomous R2-based robotaxis contingent upon Rivian achieving specific development and validation targets, with the potential to expand to 50,000 units, per SEC filings and corporate announcements. As part of the transaction, Uber is investing $300 million in the firm, with a possible additional $950 million if Rivian satisfies certain, unrevealed development criteria. Furthermore, the SEC filing indicates that Rivian has committed to withholding fully autonomous vehicles from Uber’s direct ride-hailing competitors for a defined exclusivity duration.

The companies stated that Uber aims to launch the new fleet in San Francisco and Miami in 2028, with aspirations to operate in 25 cities by 2031. Uber indicated in January that it remains on track to deploy Lucid vehicles later this year.

It is important to note that significant milestones remain before the Rivian agreement can be realized. In December, Rivian detailed the architecture of its R2 autonomy platform—comprising a multi-modal sensor suite with 11 cameras, five radars, and one LiDAR powered by two of Rivian’s proprietary RAP1 chips—however, development is not yet complete, nor has production of the new vehicle commenced. Phrasing within the SEC filings implies that Rivian still faces a considerable journey, stating that Rivian “intends to develop” an autonomous driving system incorporating its own Level 4 capabilities alongside “certain technology” to facilitate the integration of Rivian vehicles into ride-hailing and logistics networks. Additionally, the agreement reveals that Rivian and its vendors still apparently need to acquire the tooling required to manufacture and assemble these automobiles.

These endeavors will come at a high cost. As initially reported by TechCrunch, Rivian disclosed in an SEC filing that it no longer expects to achieve EBITDA profitability by 2027, citing a projected rise in autonomy research and development expenditures. It appears that Rivian may be leveraging Uber’s order commitments and capital to effectively fund this autonomous driving initiative.

Prior to publication, neither Uber nor Rivian had replied to inquiries seeking comment.

Uber shifts from an asset-light model

Uber has long placed strategic and costly wagers on autonomy, establishing itself as a leader in collaborating with numerous robotaxi firms.

The company is currently operating in cities such as Austin with Waymo vehicles and in Las Vegas with Motional, while also planning an expansion into Los Angeles with Zoox. Direct collaboration with Nvidia is also slated for 2027. These alliances have enabled Uber to maintain a vested interest in the autonomous driving race while simultaneously diverting some degree of brand and reputational liability.

Uber has compelling reasons for this approach. In 2018, a self-driving test vehicle operated by Uber struck and killed a pedestrian. This incident, the first fatal accident involving a self-driving car, caused a major public outcry. Following the suspension of Uber’s testing privileges by Arizona’s governor in 2018, the company terminated its Arizona AV program and subsequently divested it in 2020 in exchange for equity in Aurora Innovation, a Texas-based self-driving trucking enterprise.

For years, amid the signing of new partnerships, CEO Dara Khosrowshahi has maintained that Uber operates as an asset-light marketplace that does not possess its own vehicles.

These recent agreements signify a strategic pivot.

Uber is not manufacturing vehicles or developing core autonomy software—tasks handled by Rivian and Lucid—however, the company will now possess thousands of highly specialized automobiles in designated cities. This implies that Uber is assuming asset-related risks, such as depreciation and utilization rates, as well as operational liabilities in the event that these systems underperform or cause an accident.

It remains uncertain whether these new arrangements are affecting Uber’s established partnerships, or if firms like Waymo will begin to perceive Uber as a competitor. Requests for comment sent to Waymo and WeRide were not answered prior to the deadline.

For an organization that has spent years asserting that it is merely a marketplace rather than a fleet operator, this move represents more than a minor adjustment to its business model. It is a wager that, on this occasion, owning the robots will prove less detrimental than it did during the previous attempt.