Who owns whom: the deals behind robotaxis

Amazon bought Zoox, the purpose-built robotaxi with no steering wheel, and runs it as a standalone subsidiary. It's why a retail-and-cloud giant is now a robotaxi operator in Las Vegas and San Francisco.

GM acquired Cruise in 2016, then in December 2024 stopped funding its robotaxi, folded the team into GM, and refocused on personal-car driver assistance. One of the best-funded robotaxi players simply left the race.

Motional began as a 50/50 venture between Hyundai and Aptiv. In 2024 Aptiv cut its common-equity stake from 50% to 15% and said it wouldn't invest more, leaving Hyundai firmly in control.

Aurora took over Uber's self-driving group (ATG); Uber invested $400M for roughly a quarter of Aurora and left the hard AV work to a specialist. Aurora now focuses on driverless trucking.

Toyota's Woven Planet acquired Lyft's Level 5 unit. So, like Uber, Lyft exited building its own AV and became a partner that resells others' cars.

Google's self-driving project became Waymo, an independent Alphabet company, in 2016. Alphabet still controls it, with outside investors holding minority stakes from later funding rounds. It isn't separately listed.

Not acquisitions but ownership milestones: WeRide (October 2024, $15.50 per ADS) and Pony.ai (November 2024, $13.00 per ADS) became the first Chinese robotaxi makers to list in the US, with founders keeping control through multi-class shares.

Uber, the app that resells robotaxis in several cities, has built a roughly 19.5% stake (plus options) in Germany's Delivery Hero and floated an indicative offer near €33 a share, valuing it around €10B; in July 2026 it paused its Uber Eats Europe expansion to focus on it. This is a food-delivery consolidation, not a robotaxi deal, but it shows Uber building a mobility-and-delivery super-app around the AV rides it already aggregates. Note it's an indicative, non-binding approach disclosed by Delivery Hero, not an agreed deal, and Uber has not filed it with the SEC. The €33 figure is reported, not printed in a filing we could read.

Rather than buy a robotaxi maker, Uber has struck large fleet-supply deals, reportedly with Lucid and Nuro and with Rivian, for tens of thousands of autonomous cars to put on its platform. It's the asset-light aggregator playbook: resell everyone's rides, own none of the hard tech.

Both Uber and Lyft already sold their AV units and now resell others'. Buying a robotaxi maker outright is capital-heavy and cuts against that asset-light model, so more fleet deals look likelier than an acquisition.

Owning the software and the factory together is the integrated bet, and it has happened three times already (GM and Cruise, Hyundai and Motional, Amazon and Zoox). A carmaker absorbing a struggling AV startup could repeat it.

With Pony.ai and WeRide public and Baidu at scale, weaker Chinese players could be bought or fold as the crowded home market thins out.