Regulators Probe Tesla's Cybercab on Day One

Tesla began charging passengers for rides in its steering-wheel-free Cybercab in Austin this week, and federal regulators responded within a day. On September 3 the National Highway Traffic Safety Administration opened Audit Query AQ26002, examining the process and the technical data Tesla relied on when it certified the vehicle as compliant with federal safety rules.

The question at the centre of the probe is basic and hard. Roughly 1,000 Cybercabs are now on the road with no permanently attached manual controls: no steering wheel, no brake pedal, no accelerator, no side mirrors. Federal Motor Vehicle Safety Standards were written on the assumption that a human sits behind the wheel, reaches a pedal, and glances at a mirror. Regulators want to see the legal and engineering basis on which Tesla declared a car without any of those things road-legal.

An audit, not a recall

The opening documents stop short of forcing anything. As TechCrunch reported, the audit orders no recall and no halt to service, and Tesla's Robotaxi app kept accepting bookings after the announcement. What NHTSA has done is demand Tesla's homework: the certification paperwork, the test data, and the reasoning behind the company's compliance claims.

That distinction matters, because an audit query can still grow teeth. Electrek noted that the agency has used the same kind of certification review before, and that earlier robotaxi efforts have spent years working through similar scrutiny. The Cybercab has only been carrying paying riders for days.

The launch has not been quiet. During the Austin pilot, video circulated of Tesla robotaxis driving erratically in traffic, the sort of footage that tends to shorten a regulator's patience. Tesla has long argued that removing the human is the safety feature, not the risk. NHTSA's audit will test whether the paperwork supports that claim, or whether a car designed around the absence of a driver has outrun the rulebook meant to keep drivers safe.

The gap the rules never anticipated

Underneath the specifics sits a larger problem for autonomous driving. Safety standards assume the thing they regulate. When the thing changes shape entirely, the standard can end up describing a vehicle that no longer exists. This is the same friction now surfacing across AI oversight more broadly, from state legislatures rewriting the rules for algorithms to federal agencies deciding how far their old authority stretches. For Tesla, the near-term reading is narrow. The Cybercab keeps running for now, and the burden has shifted to Austin, Texas, and a stack of certification files that regulators intend to read closely.