inDrive Taps Pony.ai and Nvidia for 2027 Robotaxi Tests in Kazakhstan

inDrive has signed a memorandum of understanding with Chinese autonomous driving firm Pony.ai and chipmaker Nvidia to begin testing robotaxis in Kazakhstan's capital, Astana, in the first half of 2027, according to TechCrunch. The ride-hailing company, which built its business around letting drivers and passengers negotiate fares, is positioning the deal as a capital-light entry into autonomous mobility rather than a ground-up build.

What Happened

Under the proposed arrangement reported by TechCrunch, Pony.ai would supply its Gen-7 robotaxis, inDrive would operate them through its existing ride-hailing app, and Nvidia would provide the underlying AI stack, including its Hyperion platform and Halos safety system. Testing will start with a safety operator aboard, with fully driverless rides as the stated eventual goal. inDrive CTO Yuri Misnik told TechCrunch the approach is deliberately asset-light: "We work with a range of partners in different markets, combining their expertise with ours to deliver a better customer experience."

Nvidia said its Hyperion platform could lower robotaxi development costs by providing a ready-made foundation of sensors and compute; Ali Kani, Nvidia's vice president of automotive, separately told TechCrunch those savings are particularly relevant in emerging markets where ride-hailing is already affordable and autonomous services need to compete on cost. The announcement landed alongside Nvidia CEO Jensen Huang's first visit to Kazakhstan.

Why It Matters

The economics behind this deal are as significant as the geography. inDrive and Nvidia are explicitly framing autonomous fleets as a cost play in markets where ride-hailing fares are already thin, which points toward driver labor being the biggest line item these partnerships are trying to compress. If Hyperion genuinely lowers the capital cost of deploying a robotaxi, the productivity math shifts from paying a human driver per trip to amortizing a fixed sensor-and-compute stack across thousands of rides, a shift that doesn't show up in headline fare prices but reshapes who captures the margin. inDrive's pattern of stacking multiple regional autonomous partnerships, rather than building one global fleet, suggests the industry still sees driver displacement as a market-by-market negotiation rather than a single global transition.

Background

What to Watch

The first concrete test will be whether Astana's trials, slated for the first half of 2027, actually proceed on schedule given Kazakhstan's nascent regulatory framework for autonomous vehicles. It's also worth watching whether inDrive's Riyadh pilot with AI Driver and WeRide, originally targeted for the first half of 2026, materializes at all, since TechCrunch could not confirm its status. If either market produces data on driver earnings or fleet utilization during the transition to safety-operator and eventually driverless service, that would offer the first real signal of how wage structures shift as these deployments scale beyond pilot programs.