Tesla releases steering-wheel-free Cybercab in Austin

To put it plainly: Austin has already had Tesla robotaxis running since June of last year, and those vehicles did have a steering wheel and brake pedal. What's new here is that the Cybercab strips those controls out entirely, meaning passengers have no way to intervene themselves if something goes wrong — a key difference from the earlier service.

This is a different vehicle from last year's robotaxi

Tesla's robotaxis have been operating in Austin since June of last year, complete with a steering wheel and brake pedal. The service initially ran with a safety monitor on board who could step in immediately if something went wrong, and it later expanded to five other cities across Texas and Florida. The Cybercab that just debuted removes those controls entirely. According to Robotaxi Tracker figures, Tesla now has more than 200 "unsupervised" robotaxis — vehicles running without a safety monitor — deployed across these cities. Musk's stated plan is to push full self-driving software out as an update to hundreds of thousands of Tesla vehicles, eventually letting individual owners rent out their own cars as taxis. When the Cybercab service might expand to other cities, though, hasn't been decided yet.

The gap with Waymo, in numbers

Tesla is playing catch-up with Waymo in the driverless taxi race, trailing on both fleet size and completed rides. Waymo runs more than 4,000 unsupervised vehicles across 14 cities — nearly 20 times Tesla's roughly 200. The two companies also take different technical approaches: Tesla relies on cameras alone to perceive its surroundings, while Waymo and Amazon's Zoox add radar and lidar on top of cameras. Lidar measures distance precisely by firing light and timing how long it takes to bounce back. For Tesla to close the gap, it will need to prove that cameras alone can navigate roads and avoid pedestrians as safely as rivals that use lidar.

Trust is the real problem, and the numbers show it

Even if the technology works flawlessly, Tesla still has to win people over. In a February survey by Pew Research Center, 7 in 10 US adults said they'd be "not too" or "not at all" comfortable riding in a driverless car. Only 16% said they'd be "somewhat" comfortable, and just 7% said "very" or "extremely" comfortable. A separate Gallup survey from 2025 found that skepticism about the safety of driverless cars has grown compared to a 2018 survey. Compounding the challenge, a January Pew survey found nearly 6 in 10 US adults hold a "very" or "mostly" unfavorable view of Musk. Tesla has already weathered store protests and boycotts in multiple countries over the past year after Musk led the Trump administration's government cost-cutting effort (DOGE) and backed far-right political candidates; vehicle sales have fallen for two straight years, and the company lost its title as the world's top electric vehicle seller to China's BYD.

Regulators still haven't fully signed off

US and European regulators remain cautious. In most EU countries, even Tesla's partial self-driving software hasn't been approved for driver use. In the US, the software is downloadable and in active use, but authorities are running multiple safety investigations. One is examining crashes — including a pedestrian fatality — that occurred in low-visibility conditions like fog or bright sunlight. Another is looking at dozens of cases in which Tesla vehicles running the partial self-driving software ran red lights or drove into oncoming traffic, causing collisions and injuries. A third is examining why Tesla appears to have violated rules requiring it to report crashes in a timely manner.

Editor's Take

What stands out here is that Musk doesn't have much room to stall. Waymo is already running a fleet of more than 4,000 vehicles across 14 cities, while Tesla has only just put steering-wheel-free vehicles on the road at roughly a twentieth of that scale. Closing that gap means proving with data that camera-only perception is as safe as lidar-equipped rivals — and right now, the three ongoing investigations are, if anything, building a case for the opposite.

Anyone who's watched autonomous-vehicle rollouts before will recognize this pattern. Companies tend to lead with claims about how mature the technology is, and public opinion always lags behind. Waymo itself expanded city by city with far more caution in its early days; Tesla, by contrast, is moving faster, having gone straight to deploying multiple vehicles that lack a steering wheel altogether.

For teams here watching this case, the lesson worth taking is that regulation and public opinion don't move as fast as the technology does. Any team building autonomous vehicle or robotics services should nail down, before worrying about technical polish, exactly who is responsible and how incidents get reported when something goes wrong — that clarity matters more later than it seems now. Tesla's third investigation, over allegedly unreported crashes, is a direct product of a gap in exactly that kind of process.

Over the coming weeks, the things to watch will be how quickly the Cybercab expands to other cities and whether its early runs stay accident-free. If even one visible incident happens early on, it's likely to push already-low public trust even lower.