Uber is abandoning its 'asset-light' gig-economy model and committing over $10bn — more than $2.5bn in equity stakes and $7.5bn in vehicle purchases — to buy self-driving cars and take positions in their makers. It has announced partnerships with 12+ providers including Baidu, Rivian, Lucid, Waabi and Zoox, and plans to launch robotaxi services in at least 15 cities in 2026. Its shares are down 23% over six months as Waymo, Tesla and Zoox threaten to bypass the platform.
Uber commits $10bn to robotaxis in a strategy U-turn
— Summary
The story in one line: Uber is abandoning its famous “asset-light” model and committing over $10 bn to buy self-driving cars and take stakes in their makers, trying to stay relevant as Waymo, Tesla and Amazon’s Zoox threaten to bypass it entirely.
Key numbers
- Total commitment: >$10 bn — >$2.5 bn in equity stakes in AV developers, >$7.5 bn in vehicle purchases.
- Partnerships with 12+ robotaxi providers (including Baidu, Rivian, Lucid, Waabi, Wayve, Zoox). Robotaxi services planned in ≥15 cities in 2026.
- Lucid deal: $500 mn equity investment + at least 35,000 vehicles (~$2 bn+).
- Uber’s free cash flow last year: $9.8 bn — the AV spend exceeds a full year of cash generation.
- Uber share price: –23% in the last 6 months; each Waymo/Tesla/Zoox milestone has pushed it lower.
- Waymo’s US rideshare market share (in the cities it operates): ~7%; in San Francisco already ~16% vs. Uber’s 62%. JPMorgan forecasts Waymo at >7% of the US rideshare market by 2030.
- Uber’s “take rate” (share of each fare it keeps): 23% (end-2020) → 30% (end-2025); the US figure is <20% net of insurance and incentives.
Why it matters (jargon, explained)
- “Asset-light” = Uber owns no cars; drivers supply the vehicles. A “robotaxi” is a self-driving taxi with no human driver. By buying cars and taking equity in AV makers, Uber is becoming capital-intensive — i.e. it needs real cash to own fleets.
- “Financialisation” of the fleet: Uber plans for fleet-management firms, institutional investors and private credit funds to ultimately finance these cars — effectively turning robotaxis into a tradable asset class, the way data centres are financed today.
- The strategic bet: if Uber becomes the distribution channel (the app where riders hail the car), AV makers will still need it. The risk: giants like Waymo (Alphabet), Tesla and Zoox (Amazon) go direct to consumer with their own apps — and cut Uber out. Waymo is already doing this in San Francisco and plans the same in London.
Takeaway
Uber is defending its crown with money, not just software. A $10 bn bet that it can stay the consumer gateway for self-driving rides — against three of the world’s most valuable companies.
Source: Financial Times, 15 April 2026 — Rafe Rosner-Uddin and Tim Bradshaw.
— Delfineo's Take
The real bet is on 'financialisation': Uber plans to have fleet managers, institutional investors and private credit ultimately own and finance the robotaxis — turning self-driving fleets into a tradable asset class, like data centres. If that works, Uber stays the consumer gateway without bearing the full capital cost. If it doesn't, Uber becomes a structurally capital-intensive business competing against three trillion-dollar balance sheets (Alphabet, Tesla, Amazon). Watch the take rate — already down from 30% to under 20% net of insurance in the US.