Waymo's Phoenix exit was not a tidy end to a pilot. It was the first visible break in a partnership that now looks too small for the robotaxi market both companies want.
The breakup looked quiet only because neither company wanted to make it loud. Phoenix riders noticed Waymo cars had disappeared from Uber's app, and TechCrunch reported on June 29 that the companies had ended a nearly three-year Arizona deployment after what Uber called the contracted end date. Reuters also reported the split, with Waymo saying the vehicles used for the Uber pilot had been folded back into its own Phoenix fleet.
That is the shift.
When Waymo first worked with Uber in Phoenix in 2023, the arrangement made sense. Waymo had the self-driving system. Uber had riders already opening its app, plus the dispatch habits and customer relationships that take years to build. If you're still trying to prove that people will sit in a car with no driver, borrowing Uber's demand is useful. Once you're serving roughly 500,000 paid rides a week, as TechCrunch reported in March from Waymo's own disclosure, the math changes.
Waymo doesn't need Uber in the same way anymore. TechCrunch reported that Waymo operates across 11 U.S. metro areas and is still available through Uber in Austin and Atlanta, while The Information has reported that Waymo projects one million weekly rides by late 2026 and last raised money at a $45 billion valuation in 2024. The Information also reported that Waymo has discussed a new raise at more than $100 billion. Those figures don't make Waymo profitable. They do make it too large to behave like a supplier inside someone else's marketplace forever.
The partnership is becoming a constraint
The latest development is sharper than the Phoenix exit alone. The Financial Times reported on July 24 that Waymo is considering ending its wider Uber partnership as tensions deepen, and that Waymo has notified Uber of plans to launch independently in Austin and Atlanta from January 2028, when the current contract allows it. That doesn't mean the two companies are finished tomorrow. It does mean the end date is now part of the strategy.
Look at the incentives. Uber wants to stay the front door for urban transport, whether a ride is driven by a person, a Waymo vehicle, a Wayve system, or another partner. Waymo wants the customer relationship for itself, because the app is where pricing, loyalty, routing data, and brand trust live. A robotaxi company that wins the road but rents the customer from Uber has not really won the business.
The relationship has also become awkward in public. In April, Uber CTO Praveen Neppalli posted a video on X showing what he described as a scary Waymo moment in San Francisco, where a robotaxi overtook a Muni bus in the wrong lane and tried to squeeze between his vehicle and the bus. Yahoo Finance captured the post. That's not normal partner behavior. You don't publicly call out a key partner's vehicle unless the partnership is already under strain.
Uber is fighting the rules now
The cleaner fight is in regulation. TechCrunch reported on July 13 that Uber has opposed a Washington, D.C. bill that would allow fully driverless commercial robotaxi operations, with a $1 million application fee, a $5 million permit fee, at least $5 million in liability insurance, and a 15-cent-per-mile tax split between public transit and workforce programs. Waymo backs the bill. Uber argues it could hand Waymo a de facto monopoly and hurt human for-hire drivers.
Read Uber's alternative twice. TechCrunch reported that Uber has pushed for a hybrid model requiring robotaxis to operate on networks that also include human drivers. In New Jersey, Wired reported on lobbying language that would require human drivers to complete 85 percent of rides on any platform offering driverless service for three years. That would not merely regulate Waymo. It would push Waymo back toward platforms like Uber.
That's the real issue. Safety is part of every autonomous vehicle debate, and it should be. But this proposal is also a market structure argument. Uber is trying to make the platform the required layer. Waymo is trying to prove the vehicle operator can also own the rider relationship.
San Francisco has given regulators fresh material. Local outlets including the San Francisco Chronicle, the San Francisco Standard, NBC Bay Area, and ABC7 reported that dozens of Waymo vehicles were caught in July 4 gridlock around the Presidio and Crissy Field after the Golden Gate Bridge fireworks, with some vehicles running down their batteries and needing to be towed. CBS San Francisco reported that one unoccupied Waymo caught fire after driving over a small firework, with no injuries reported. The Chronicle also reported that Supervisor Bilal Mahmood sent formal inquiries to city agencies, and Mayor Daniel Lurie later asked California officials for tougher robotaxi standards during major events and emergencies.
None of that kills Waymo's lead. It does make the next phase harder. The company is no longer judged as a science project or a novelty ride. It's being judged as transport infrastructure, which means dead batteries, blocked streets, disabled vehicles, and emergency planning count as much as smooth rides on ordinary afternoons.
For investors watching autonomous vehicle startups, the lesson is uncomfortable but plain. The intermediary model was a bridge, not the destination. If Waymo keeps scaling its own app while Uber tries to write hybrid networks into law, the old partnership doesn't fade gently. It turns into the opening fight over who owns the robotaxi customer.
Also read: Morgan Stanley says SpaceX at $100 prices its entire AI business at zero • Google Zero is here and every startup built on SEO traffic just got a hard lesson • Cognition buys Poke to plant its AI agent inside your text messages
