Tesla Robotaxis Accused of Speeding in School Zones
Federal auto safety officials are checking claims that Tesla robotaxis drove too fast in active school zones.
The case centers on a safety petition sent to NHTSA's Office of Defects Investigation, the office that checks for car defects, on September 23, 2026. It asked for an investigation into Tesla Robotaxis The Verge.
The scope is narrow. It covers 2026 Model Y and Cybercab vehicles when the self-driving system, the software that drives the car, is engaged.
The petition says the system goes over posted limits when school zones are active. NHTSA has published a short receipt document, INOA-DP26010-19159, for those vehicles NHTSA. That publication does not open a formal investigation. It only records that the petition was received and is being processed.
Separate coverage describes a Tesla robotaxi going 40 to 45 mph in a 35 mph zone, seen by a Reuters reporter Yahoo Finance. That report alone does not prove the petition claim. Concerns about robotaxi speeding were reported as early as 2025, according to the same coverage.
The broader context here explains why school zones are hard. They use lower limits at certain times, flashing lights and signs that vary by location. A person can read the scene quickly. A self-driving system must spot each sign, decide if the lower limit applies and choose a speed within the conditions it was built for, while keeping records that show what it saw and why. It is like a student driver who must learn that a flashing light means slow down right now.
Looking at what this means for safety checks, the petition process is what to watch. A defect petition asks officials to decide if a safety defect exists and if federal action is needed. Review usually looks at what happened, how often, under what conditions and what data is needed to judge risk. Speeding in a school zone draws strong concern because children are present and full compliance is expected when the zone is active.
In my view, the key question is narrow. Does the system reliably notice when the lower school-zone limit is on and treat it as more important than other inputs such as old map data, its last speed or the speed of traffic. That priority problem is common in self-driving systems and in driver-assist speed tools. Being extra careful can cause unneeded slowdowns that annoy riders, but the other choice is breaking a limit where mistakes carry high cost.
Looking ahead, the answer will turn on data. Regulators will need trip logs, sensor records and the logic used to pick the speed limit for that software version, not stories. Operators will need to show they can obey the limit each time across different zone types and on and off states. The most useful outcome would be a clearer rule for how companies test and document compliance with these changing limits.
