Tesla Cybercab Hidden Joystick: What It Means for Safety
On September 7, a Tesla Cybercab rider named Matthew Skrzypczak climbed into what was supposed to be a fully autonomous robotaxi and found something Tesla has never publicly acknowledged: a virtual joystick on the touchscreen, paired with a live surround-camera feed, door controls, and a button to call support. He posted the video to X. He was not the only one. David Moss posted a similar clip a day later, describing the same unexpected interface from what appeared to be a different vehicle. Neither rider could get the car to move — but that’s almost beside the point.
The discovery lands at the worst possible moment for Tesla‘s autonomy narrative. The Cybercab launched commercially in Austin on September 3, and NHTSA opened Audit Query AQ26002 the same day, examining how Tesla self-certified a vehicle with no steering wheel, no pedals, and no mirrors under federal safety standards written for traditional cars. Now, four days after that audit opened, there is documented evidence of a manual control interface that Tesla never disclosed to regulators, never described in its public rider guides, and apparently never secured against passenger access.
What the Joystick Actually Is — and Why That Raises More Questions
Tesla’s first responder documentation does acknowledge that the Cybercab can be manually moved using onboard controls. According to that guide, emergency personnel are supposed to call Tesla support and show their badge to a Cybercab’s cameras before gaining access. That process implies the joystick mode is real, intentional, and gated — a legitimate service or emergency tool, not a software accident.
But Skrzypczak didn’t call anyone or flash a badge. He just got in. The interface was already on the screen, apparently left active from a prior Supercharger stop — both riders noted that Supercharger stalls were visible in the rear camera feeds, suggesting the manual mode had been used to reposition a vehicle and was never properly exited. The car was then dispatched to a paying passenger with that interface still live.
The controls themselves are straightforward: slide the joystick forward to move forward, drag left or right to steer, up to a reported maximum of 7 mph. Whether any additional authentication step was required to actually move the vehicle — and whether that step was present or absent in Skrzypczak’s case — Tesla has not explained. That silence is the core of the problem.
The Regulatory Gap NHTSA Didn’t Know It Had to Ask About
Tesla’s self-certification strategy for the Cybercab rested on a specific argument: that federal safety standards covering steering columns, brake pedals, and mirrors simply don’t apply to a vehicle engineered without those components. NHTSA is auditing whether that reasoning holds. What the agency was not auditing — because it had no reason to — is whether a manual control interface exists inside the car and what access controls govern it.
That gap matters. Tesla chair Robyn Denholm acknowledged last October that the Cybercab could be equipped with a steering wheel and pedals if regulators required it. That comment read, at the time, as a contingency. It reads differently now that a touchscreen driving interface has surfaced in a passenger-facing cabin. The fallback was always there. It just wasn’t disclosed.
Amazon’s Zoox took a different path with its own steering-wheel-free shuttle: it self-certified first, then filed for a formal Part 555 exemption, which NHTSA approved in July 2026 with explicit documentation of every standard from which the vehicle was excused. That process created a public record. Tesla’s self-certification approach created no equivalent transparency — and the joystick discovery suggests there were relevant details that never entered the regulatory conversation.
The Trust Problem That Now Sits in Every Cybercab
Tesla’s public documentation for Cybercab riders covers passenger conduct, emergency stops, and how first responders should approach a stranded vehicle. It does not mention a virtual joystick. It does not describe any scenario in which a rider might encounter a manual driving interface, or what they should do if they do. That gap is not a minor omission — it is the difference between a safety feature and a liability.
The practical question is straightforward: what prevents a passenger from finding this menu and attempting to use it? The answer, based on everything Tesla has made public, is nothing documented. Skrzypczak tried. The car didn’t respond — but whether that was because of an authentication requirement, a software lock, or simply because the vehicle was stationary at a Supercharger, nobody outside Tesla knows.
The Cybercab was pitched as proof that full autonomy has arrived and that manual controls are obsolete. Prototypes were spotted with conventional steering wheels during testing; the production vehicle deleted them to make a statement. That statement now has an asterisk. A car with no wheel and no pedals still has a way to be driven — it’s just buried in a menu, undocumented, and apparently accessible to whoever happens to get in at the wrong moment. Tesla owes riders, regulators, and the cities where it plans to expand a clear answer to the question Skrzypczak’s video raised: who gets to drive it, and what’s actually stopping anyone else?
Related posts
Latest Post
No Comments