NHTSA Opens Formal Investigation into Cybercab's Road Deployment
nashnova research
NHTSA announced a formal investigation into Tesla's Cybercab just hours after the first units hit Austin streets — the regulator moved on its own rather than waiting for Tesla to apply for an exemption, and the core bottleneck for scale-up is now official.
Why did NHTSA act at this exact moment?
On September 5, NHTSA formally opened an investigation into Cybercab's public-road operations — just hours after the first vehicles drove onto Austin streets.
This means → the regulator did not wait for Tesla to file an exemption request. It stepped in proactively, taking control of the timeline.
NHTSA had previously been in an "assessment" phase. The upgrade to a formal investigation turns an abstract regulatory risk into a concrete obstacle.
Where does Cybercab's design clash with federal law?
Cybercab has no steering wheel, no pedals, and no mirrors. Current Federal Motor Vehicle Safety Standards (FMVSS) explicitly require manual controls.
Tesla has "self-certified" to NHTSA that Cybercab meets all standards — self-certification is standard industry practice, but a car with no steering wheel claiming full compliance is precisely the point of contention.
In plain terms = the U.S. Department of Transportation has proposed exempting purpose-built autonomous vehicles from these requirements, but the proposal is not yet in effect. The rules haven't changed; the car is already on the road.
How small is the fleet, really?
Texas currently has 45 registered Cybercabs for paid autonomous rides, plus roughly 375 Robotaxi-variant Model Ys — 420 vehicles in total.
For comparison: Waymo operates about 4,000 vehicles across 14 U.S. cities.
Goldman Sachs cautioned that not all 45 Cybercabs may be in commercial service — some may still be in testing or staffed with safety operators. This means → the actual commercial fleet could be even smaller than the headline number.
Is the cost advantage enough to support the business case?
Goldman estimates that at scale, Cybercab could cost $20,000–$30,000 per unit, versus $50,000–$100,000 for some competitors — yielding roughly $0.05–$0.30 per mile in cost advantage.
But both Goldman and Barclays flagged the same point: what determines Robotaxi economics in the near-to-medium term is not vehicle cost but whether the self-driving software can support expansion across cities.
In plain terms = a cheap car is one thing; software that works reliably in city after city is another — the latter is the real bottleneck.
What can the safety data actually tell us right now?
Goldman, drawing on NHTSA collision data from January to mid-July 2026, estimates Tesla's Robotaxi fleet averages one incident per 50,000–70,000 miles driven (regardless of fault).
But Tesla only launched fully driverless passenger service in January. The fleet is still small, and the existing data sample is insufficient to assess safety at scale.
This reflects a fundamental catch-22: proving safety requires large-scale data, but large-scale operations require regulatory clearance first — data and permission are each the other's prerequisite.
Can Musk's timeline still hold?
Musk previously promised investors: Robotaxi service in half of U.S. cities by end of 2025, millions of vehicles on the road by end of 2026, and a Cybercab price tag under $30,000.
Tesla has built annual production capacity exceeding 125,000 Cybercabs — capacity is not the constraint.
This means → everything now hinges on whether NHTSA's investigation can close quickly and when exemption rules take effect. The factory is ready; the regulatory gate is not open.
市场有风险，内容仅供研究参考，不构成投资建议。