Waymo has ended its three-year robotaxi partnership with Uber in Phoenix, folding its vehicles back into its own fleet and leaving Uber scrambling for a replacement partner it won't yet name.
The breakup was quiet enough that riders noticed before either company said anything. Waymo's vehicles simply disappeared from Uber's Phoenix app sometime in May, according to TechCrunch, and it took until June 29 for Bloomberg to confirm what had happened: the arrangement that began in 2023, covering both passenger rides and on-demand food deliveries, is over. Waymo's Phoenix fleet is now operating exclusively through Waymo's own app, including public transit connections with Via and deliveries through DoorDash. Uber says a new autonomous partner for Phoenix is coming. It hasn't said who.
That gap matters more than the breakup itself.
When Waymo signed its multiyear deal with Uber in 2023, the logic was straightforward. Uber had the demand, the driver network relationships, and the brand recognition. Waymo had the technology but not the riders. Phoenix, where Waymo launched paid passenger service in 2020, was the right test bed: a sprawling, sunny, grid-planned city that autonomous vehicles handle well. The partnership let Waymo fill cars without building its own customer acquisition machine from scratch.
Two and a half years later, Waymo doesn't need that machine. The company is now providing over 400,000 paid rides per week across its commercial markets, according to Road to Autonomy, with a fleet of roughly 2,500 vehicles operating across ten metro areas. It has announced expansions into Dallas, Houston, San Antonio, Orlando, Chicago, Charlotte, London, and Tokyo, with a target of 1 million weekly rides and a plan to quintuple its city count from five to twenty-five. Notably, as CNBC reported in February when Waymo opened service to select riders in four new cities, the company's expansion announcements have made no mention of Uber as a distribution channel. Waymo is building its own funnel now. In Nashville, it signed with Lyft, Uber's largest domestic rival. In London, where the two companies are headed for direct competition, Waymo went with fleet operator Moove rather than Uber.
Waymo's vehicles are still available on Uber in Austin and Atlanta, so this isn't a total divorce. But Phoenix was where the relationship started, and losing it there is symbolic in a way that Austin and Atlanta can't quite offset. The city is the proof of concept for the whole distribution partnership model, and Waymo just decided it can do without the model in the market where it has the most experience.
Uber's position is more complicated. The company has been pouring capital into building an autonomous vehicle portfolio that doesn't depend on any single partner, and that strategy predates the Phoenix split by months. In March, Fortune reported that Uber struck a deal for up to 50,000 Rivian-built robotaxis, with an investment of up to $1.25 billion, targeting San Francisco and Miami starting in 2028. Uber has Avride lined up for Dallas and Wayve for London. It launched Uber Autonomous Solutions as a dedicated business unit. On paper, Uber is hedging smartly across a diversified AV supplier base rather than being held hostage to Waymo's pricing and terms.
But here's the thing: none of those partners are Waymo. Avride, Wayve, and a Rivian robotaxi that doesn't exist yet are not equivalent replacements for the company that has logged more autonomous miles than any competitor and is already putting 400,000 people in driverless cars every week. When Uber announces its mystery Phoenix partner, the first question will be whether that company can actually deliver rides at scale today, or whether Uber is plugging a gap with a promise.
The deeper issue is unit economics. Waymo's decision to pull back its Phoenix fleet and serve riders directly means it captures the full fare rather than splitting revenue with Uber. As Waymo scales, that margin difference compounds. Waymo was always the better product in Phoenix. Now it's also the only place to get it there. Riders who want a Waymo in Phoenix have to use the Waymo app, which is precisely where Waymo wants them. Every autonomous mile Waymo drives on its own platform is a mile that trains its pricing power, its brand loyalty, and its data flywheel without sharing any of it downstream.
Uber built its entire business on being the layer between riders and vehicles, taking a cut regardless of who owns the car or drives it. That model works beautifully when the supply side is fragmented human drivers. It gets harder when the supply side is a vertically integrated technology company with $5 billion in Alphabet backing that has decided it would rather own the customer relationship itself. Waymo isn't just a better robotaxi. It's a competitor that spent three years learning Uber's Phoenix demand patterns from inside Uber's own network. The partnership was always going to end this way once Waymo had enough scale to walk. It just ended sooner than most expected.
Also read: London Stock Exchange Group turns AI from threat into its strongest growth engine • Congress is circling the health data business model that AI startups built their futures on • Taiwan Raided Super Micro's Offices Today as the Nvidia Chip Smuggling Case Reaches a New Front
