Uber is trying to sell what it valued at $7.25 billion just eighteen months earlier
In the restless churn of technological ambition, Uber finds itself releasing what it once prized most — its self-driving division, valued at $7.25 billion just eighteen months ago, now quietly shopped to Aurora Innovation as the ride-hailing giant retreats toward its profitable core. The potential sale reflects a broader reckoning across the autonomous vehicle industry, where the distance between vision and viability has proven longer and costlier than anyone anticipated. Aurora, built by veterans who once shaped the very programs Uber sought to rival, may inherit not just the talent and technology, but the unfinished promise of a future that has yet to arrive.
- Uber's self-driving unit, once the crown jewel of its moonshot ambitions, has become a $303 million annual drain that the company can no longer justify as it fights for post-pandemic survival.
- Aurora Innovation — founded by alumni of Google, Tesla, and Uber itself — is in advanced talks to absorb a division that would more than double its workforce overnight and hand it a coveted Toyota partnership.
- The deal carries the weight of a troubled history: a fatal crash in Arizona, a $244.8 million settlement with Waymo over stolen trade secrets, and a valuation now sliding toward a down round.
- Structuring the acquisition is the central puzzle — Aurora's $2.5 billion valuation makes an outright purchase difficult, pushing both sides toward a creative equity-retention arrangement modeled on Uber's earlier Jump scooter divestiture.
- The autonomous vehicle sector is quietly consolidating around a handful of survivors, and this transaction, if it closes, would redraw the competitive map by merging two of the industry's most talent-dense organizations.
Eighteen months ago, Uber's self-driving division commanded a $7.25 billion valuation, freshly backed by Toyota, DENSO, and SoftBank's Vision Fund. Today, Uber is trying to sell it. Aurora Innovation — a startup led by three autonomous vehicle veterans who previously ran programs at Google, Tesla, and Uber — is in advanced talks to acquire the unit, with negotiations progressing since October toward what could be an imminent deal.
Uber's retreat from self-driving is part of a sweeping strategic retrenchment. After going public and absorbing the shock of the pandemic, the company shed its scooter unit, sold a stake in freight, and acquired Postmates. The self-driving division became the last expensive bet still on the table — one that lost $303 million in the first nine months of 2020 alone. Industry sources say the unit was facing a potential down round, a signal that its peak valuation had become difficult to defend.
Aurora was founded in 2017 by Sterling Anderson, Chris Urmson, and Drew Bagnell — figures who collectively shaped the autonomous vehicle ambitions of Tesla, Google, and Uber. The company has grown to 600 employees and expanded into long-haul trucking alongside its robotaxi work, but it remains smaller than Uber ATG's 1,200-person operation. Roughly 12 percent of Aurora's current workforce came from Uber, making the cultural overlap significant.
Uber's own history in autonomous vehicles is a cautionary tale. Its 2016 acquisition of Otto, a self-driving truck startup, triggered arbitration from Google and ultimately a lawsuit from Waymo alleging trade secret theft — a case that settled in 2018 for $244.8 million in equity. A fatal crash in Tempe, Arizona in March 2018 halted all testing and collapsed the company's ambitious timeline of deploying 75,000 autonomous vehicles across thirteen cities by the early 2020s.
What makes Uber ATG attractive to Aurora is not only its engineering talent but its relationship with Toyota, which had committed $500 million to Uber before the 2019 spin-out and had announced plans to deploy automated Sienna-based vehicles on Uber's network in 2021. That partnership would transfer to any buyer.
The deal's structure remains unresolved. Aurora's $2.5 billion post-money valuation makes a full acquisition a financial stretch, and observers expect the transaction may mirror Uber's Jump precedent — allowing Uber to offload operational costs while retaining an equity stake and a claim on any future upside. For Uber, it would be a way to close a costly chapter. For Aurora, it would be a chance to absorb the ambitions of an era and carry them forward.
Eighteen months ago, Uber's self-driving division was worth $7.25 billion. Toyota, DENSO, and SoftBank's Vision Fund had just poured a billion dollars into it. Now Uber is trying to sell it. Aurora Innovation, a startup founded by three autonomous vehicle veterans who previously led programs at Google, Tesla, and Uber itself, is in advanced talks to buy the unit, according to three people with knowledge of the negotiations. The two companies have been discussing terms since October, and the conversations have progressed far enough that a deal could be close.
Uber's shift away from its self-driving ambitions reflects a broader strategic retreat. Two years ago, the company operated like it wanted to own every form of transportation—ride-hailing, scooters, food delivery, freight, and eventually robotaxis. That changed after Uber went public and accelerated when the pandemic reshaped how people moved and shopped. In the past eleven months, Uber has shed its scooter unit Jump, sold a stake in its freight business, and acquired Postmates. The self-driving division has been the last major expensive bet on the table. In the nine months ending September 30, 2020, Uber ATG and its other experimental technology programs lost $303 million. The company had spent $457 million on research and development for these initiatives. Industry sources say Uber has been actively shopping the unit to multiple companies, including automakers, and that the division was facing a potential down round—a sign that its valuation was under pressure.
Aurora was founded in 2017 by Sterling Anderson, who led Tesla's Model X development and Autopilot program; Chris Urmson, who ran Google's self-driving project before it became Waymo; and Drew Bagnell, a Carnegie Mellon professor who helped launch Uber's autonomy efforts in Pittsburgh. The company has grown to 600 employees across San Francisco, Pittsburgh, Texas, and Bozeman, Montana, where it operates Blackmore, a lidar company it acquired in 2019. About 12 percent of Aurora's workforce previously worked at Uber. Despite this growth, Aurora is still smaller than Uber ATG, which employs more than 1,200 people across Pittsburgh, San Francisco, and Toronto. Uber owns 86.2 percent of ATG on a fully diluted basis.
Uber's journey into autonomous vehicles began in 2015 with a partnership with Carnegie Mellon's robotics center, which led the company to recruit dozens of researchers. A year later, Uber acquired Otto, a self-driving truck startup founded by Anthony Levandowski and three other Google engineers. The acquisition proved toxic almost immediately. Google filed arbitration demands against Levandowski and Ron within two months. Then Waymo sued Uber in February 2017, alleging that Levandowski had stolen trade secrets that made their way into Uber's systems. The lawsuit went to trial and settled in 2018, with Uber agreeing to pay $244.8 million in equity and promising not to use Waymo's confidential information. Early on, Uber had estimated it could deploy 75,000 autonomous vehicles by 2019 and operate driverless taxi services in thirteen cities by 2022, spending $20 million a month to get there. A fatal crash in Tempe, Arizona in March 2018 involving one of Uber's test vehicles halted all testing and derailed those timelines. Technical obstacles, the Waymo litigation, and the Levandowski controversy combined to make the project far more difficult than anticipated.
Uber spun out the self-driving division in spring 2019 after closing the $1 billion funding round, about a month before the company's initial public offering. The spin-out was meant to share the financial burden with other investors and let Uber focus on near-term profitability. What makes Uber ATG attractive to Aurora is not just its talent but its relationship with Toyota. The Japanese automaker had already invested $500 million in Uber before the 2019 injection, and the two companies had announced plans to deploy automated Toyota Sienna-based ridesharing vehicles on Uber's network in 2021. That partnership would transfer to any buyer of the division.
Aurora itself has had a mixed track record with partnerships. It secured deals with Hyundai, Byton, and Volkswagen Group early on, but VW ultimately concluded its partnership and backed Argo AI instead, which locked up Ford as a customer. Hyundai took a minority stake in Aurora but also formed a 50-50 joint venture with Aptiv called Motional, committing $4 billion in combined value. Aurora did raise $530 million in a Series B round last spring led by Sequoia, with significant backing from Amazon and T. Rowe Price, valuing the company at $2.5 billion post-money. The company has been expanding its focus beyond robotaxis into long-haul trucking, particularly after acquiring Blackmore. In July 2020, Aurora announced it was testing commercial routes in the Dallas-Fort Worth area with Fiat Chrysler Pacificas and Class 8 trucks.
How the deal would be structured remains unclear. Even with Uber ATG's valuation depressed from its $7.25 billion peak, acquiring it outright would stretch Aurora's resources unless the startup secured additional outside funding or structured the transaction to let Uber retain an equity stake. Uber has precedent for this approach. Earlier in 2020, it led a $170 million investment round into Lime while offloading its Jump scooter unit to the company. Industry sources say interest in selling ATG accelerated after the Jump deal closed, and one investor described the potential Aurora transaction as an attractive Plan B for Uber—a way to get the expensive division off its books while potentially capturing some upside from its future success.
Citas Notables
An interesting Plan B for Uber, a deal that would allow the company to take ATG off the books, while potentially getting to benefit from a little upside.— Industry investor quoted in reporting
La Conversación del Hearth Otra perspectiva de la historia
Why did Uber suddenly decide to sell something it valued at seven billion dollars just a year and a half earlier?
The pandemic forced a reckoning. Uber went public in 2019 and realized investors wanted to see a path to profitability, not a sprawling bet on every possible transportation mode. Self-driving cars are decades away from generating real revenue. Meanwhile, the company was burning through hundreds of millions a year on ATG alone.
So it's purely financial pressure?
Not entirely. The Waymo lawsuit, the fatal crash in Arizona, the Levandowski scandal—these all made the project harder and more expensive than Uber anticipated. By 2020, the company had already spun out the division once to share costs. Selling it entirely was the next logical step.
Why would Aurora want to buy something that's been so troubled?
Because Aurora gets three things at once: over a thousand experienced engineers, a deep relationship with Toyota that includes a commitment to deploy vehicles on Uber's network, and a chance to consolidate the fragmented autonomous vehicle industry. Aurora is still smaller than ATG. This deal would triple its size overnight.
But Aurora is much younger and smaller. How does it afford this?
That's the real question. The deal probably won't be a straight cash purchase. Uber might retain an equity stake, the way it did with Jump and Lime. Aurora might bring in new investors. Or the structure could be something hybrid—Uber takes some cash, some equity, and some ongoing revenue share.
What happens to the Toyota partnership if the deal closes?
That's the crown jewel. Toyota invested in Uber ATG specifically to develop autonomous ridesharing vehicles. If Aurora buys the division, that relationship transfers. Toyota gets a new partner with fresh energy and a cleaner slate. For Aurora, it's a customer and a deep-pocketed backer all at once.
Is this the end of Uber's self-driving ambitions?
Probably. Uber is consolidating around ride-hailing and delivery—the businesses that work today. Self-driving is someone else's problem now.