Uber's Data Hedge: Why AV Labs Matters More Than the Hotel Bookings

Uber's Chief Product Officer Sachin Kansal used a wide-ranging TechCrunch interview to sketch a company expanding well beyond ride-hailing and delivery, according to the source. Hotels via Expedia, shop-for-me concierge features, and boat rentals in Europe now sit alongside the core apps, framed by Kansal as travel becoming a third pillar after rides and eats. Yet the more consequential disclosures concerned autonomy, financial services restraint, and a data operation called AV Labs, launched six months ago according to the reporting.

For automotive leaders, the AV Labs initiative deserves particular attention. Kansal described equipping hundreds of vehicles with sensors through fleet partners to gather what he called long-tail driving data, covering edge cases beyond typical performance benchmarks. He also pointed to operational knowledge accumulated across Uber's driver base, including handling 25 million lost items annually, as a distinct asset for autonomy partners. TechCrunch's framing of AV Labs as a hedge against Waymo and other L4 providers, in which Uber holds equity while also competing with them in various cities, captures a structural tension that senior leaders in the sector should track closely. Kansal's own description of the Phoenix pilot wind-down as a mutual decision, even as Uber scales hundreds of Waymo vehicles in Austin and Atlanta, illustrates how partnership and rivalry now coexist within the same commercial relationship.

Kansal's insistence that Uber is not seeking to become an L4 autonomy provider itself, and instead wants to lay down what he termed the race tracks for multiple players, signals a deliberate strategic choice to remain platform-agnostic while accumulating data leverage. This positions Uber as an infrastructure layer across autonomy providers rather than a competitor to any single one, a stance that could prove advantageous if the hybrid network model he described, combining human drivers and autonomous vehicles within the same city to balance supply and demand, becomes the dominant deployment pattern.

On monetisation, Kansal confirmed that Uber sells earner-labelled data to generative AI companies, describing it as an emerging and promising business line, while stating that AV Labs data-sharing arrangements remain unresolved. He was explicit that driver-rider conversations during trips are not recorded, distinguishing this from separate paid transcription work drivers can undertake outside active trips. Automotive leaders should watch how Uber formalises data-sharing terms with AV partners, whether the hybrid network model gains traction elsewhere, and how quickly agentic AI features, which Kansal declined to date but described as inevitable, move from experimentation to deployment.