Replies to post #102311 on Tesla Inc (TSLA)

[zen222]

Tesla’s ‘No Fallback’ Cybercab Uses System That Crashes 4X Worse Than Humans—NHTSA Now Probing 3.2 Million Vehicles

For human drivers, insurers expect only one every 229,000 miles. Tesla’s track record is four times worse.



Tesla’s Austin Robotaxi fleet operates about 42 cars. On any given day, fewer than one in five are out on the road. In ten months of commercial service, the fleet filed five new crash reports in January 2026, covering mishaps from December and January. Federal safety regulators have started to take notice.

A crash in July 2025 was initially classified by Tesla as "property damage." Five months later, in December, records changed to include hospitalization. Tesla kept that detail undisclosed. Every Robotaxi crash report sent to regulators is redacted as "confidential business information." No other autonomous vehicle company follows this approach to the same extent. This pattern reflects legal maneuvering designed to manage liability through red tape.



Elon Musk described the Cybercab: "There's no fallback mechanism. This car either drives itself, or it does not drive." The vehicle has no steering wheel, no pedals, and no manual override. In paperwork to California regulators, Tesla confirms the current Robotaxi service requires a trained human driver in every car and support from remote operators. The present system, operating at Level 2, depends on human involvement.

Tesla has logged over 8 billion miles with its supervised Full Self-Driving system, more than any competitor. Despite this mileage, the crash rate remains four times higher than that of human drivers. The challenge is not a lack of data. The system displays brittleness in ordinary conditions: sudden phantom braking, hesitation at red lights, driving over curbs, or speeding through 30 mph zones at 37. The technology falls short on everyday tasks. Cybercab uses this same system as its foundation.



After nine crashes where Tesla’s Full Self-Driving system failed to warn drivers in poor visibility, including fog, sun glare, and dust, NHTSA expanded its investigation. The case moved to an engineering analysis, one step from a potential recall of 3.2 million Tesla vehicles. In August 2025, a Miami jury awarded more than $240 million to crash victims in the first successful jury verdict against Tesla in a fatal Autopilot case. This was not a settlement. A jury reviewed the evidence and assigned responsibility to Tesla.

Shareholders filed suit alleging Elon Musk and Tesla CFO Vaibhav Taneja made "materially false and misleading statements" about robotaxi readiness, overstating autonomous driving effectiveness and downplaying operational risks. The current crash rate has prompted some insurance companies to consider restricting Cybercab coverage, potentially forcing Tesla to self-insure. If NHTSA issues a formal recall notice before quarterly earnings, the impact on valuation could be large. The $240 million Miami verdict opened the door to further legal exposure, which had previously been limited by confidential settlements.



NHTSA's engineering analysis could conclude by end of Q2 2026, during the Cybercab production ramp. State regulators in California, Texas, and New York could impose stricter conditions than any federal exemption. Shareholder lawsuit discovery could surface internal emails revealing gaps between public autonomy promises and engineering assessments. Each new Autopilot verdict following Miami's more than $240 million precedent signals plaintiff attorneys that juries will hold Tesla liable. The legal exposure compounds with every mile driven and every crash redacted.

Tesla has not applied for the NHTSA exemption required to sell a vehicle without steering wheel or pedals. The Cybercab trademark remains contested. A French beverage company, UNIBEV, has until late April 2026 to respond to Tesla's opposition filing. Tesla may not legally own the name when mass production begins. The "driverless" Robotaxi still requires human drivers. The car without a steering wheel operates on a system that continues to underperform compared to a human behind the wheel.