Analysis
Uber and Pony.ai are scaling up. The two companies announced an expansion of their strategic partnership to deploy more than 2,000 Pony.ai robotaxis across Europe, according to CNBC, moving past their initial commercial launch in Zagreb, Croatia into four additional European cities that haven't been named yet, with further expansion planned for the Middle East.
The structure is the interesting part. Pony.ai's so-called joint-deployment model splits the three pieces a scaled robotaxi service actually needs across specialized partners: Pony.ai supplies the Level 4 autonomous driving stack, Uber supplies the rider-facing mobility platform and demand, and Verne -- the fleet operator involved since a three-way deal in March 2026 -- owns and runs the physical vehicles day to day. That's a meaningfully different model than Waymo, which under Alphabet has largely kept vehicle ownership and operations in-house, or Tesla, which is trying to run its own robotaxi network directly. Zoox, Amazon's robotaxi unit, similarly controls its own fleet.
The partnership isn't new -- Uber and Pony.ai first announced plans to put Pony.ai vehicles on the Uber platform internationally back in May 2025, and Zagreb became Europe's first commercial robotaxi service when it launched in April 2026. What's new is the scale: 2,000-plus vehicles is an order of magnitude beyond a single-city pilot, and signals both companies believe the unit economics of the joint-deployment model are working well enough to replicate quickly.
“Zoox, Amazon's robotaxi unit, similarly controls its own fleet.”
The timing also carries geopolitical weight. Pony.ai is a Guangzhou-and-Beijing-headquartered company, and Uber striking a large European expansion deal with a Chinese autonomous-driving firm comes as the US and EU both scrutinize Chinese AI and hardware partnerships more closely across other sectors. European regulators approving thousands of Chinese-technology robotaxis on public roads is a different political conversation than a software licensing deal, and it's one worth watching as the rollout proceeds -- particularly if Washington's posture toward Chinese AV technology hardens further this year.
For Uber, this continues a pattern of not building its own self-driving stack and instead stitching together partnerships -- Waymo, Nuro, and now Pony.ai -- across its autonomous vehicle empire, a strategy that lets it capture platform economics without carrying AV R&D costs directly. That approach has been tested this year: Uber's relationship with Waymo has been publicly strained by reports of a potential breakup, which Uber's own CEO has had to deny, underscoring how much leverage Uber's AV partners hold when they can credibly threaten to sell directly to riders without a platform intermediary.
Pony.ai brings a different risk profile than Waymo. It's a smaller, foreign-listed company without Alphabet's balance sheet, and its US listing has already drawn scrutiny typical of Chinese ADRs -- audit-access questions, geopolitical delisting risk, and a valuation that swings harder on regulatory headlines than Waymo's ever would as an Alphabet subsidiary. Uber is effectively diversifying its AV supplier base by adding a partner with a very different risk-and-reward profile, spreading its bets across US, Chinese and independent AV technology providers rather than depending on any single one.
Competing models are also diverging on unit economics. Tesla's robotaxi push keeps everything in-house, betting that vertical integration eventually produces the lowest cost per mile; Zoox, under Amazon, is similarly closed. Uber's multi-partner, asset-light model spreads capital risk across partners but concedes some margin to each one -- a tradeoff that will look smart if AV hardware costs stay volatile, and less smart if a single winning AV stack eventually captures most of the market's demand.
Watch for which four cities get named next and whether any EU regulator publicly objects to the scale before vehicles hit the road.
