Since ENN’s last Cybercab update on the Giga Berlin factory plan, the story has moved from where the robotaxi will be built to how it is driving on real roads. Tesla’s Cybercab now has eight crashes on file with NHTSA, barely two months after the purpose-built robotaxi started giving public rides in Austin, according to an Electrek analysis published November 16.1
The number needs its context, and the context is encouraging for the driverless program. Four of the eight happened in driverless Robotaxi service, and none of those topped 5 mph.1 The other four happened with a test driver on board and Tesla’s driver-assist engaged, including one where a Cybercab hit a car at 53 mph on a highway in Philadelphia.1 Electrek’s own conclusion is that the data points to speed as the biggest limitation of Tesla’s self-driving program right now.1
Two separate federal files
Tesla files Cybercab crashes under NHTSA’s Standing General Order in two separate datasets.1 The first is the ADS (Automated Driving System) file, which covers Level 4 driverless operation: the Robotaxi service, where nobody sits in the driver’s seat.1 NHTSA’s latest ADS release, published this month, contains 43 unique Tesla reports, up from 22 in July.1
The second is the Level 2 ADAS file, which covers driver-assist systems such as Autopilot and Full Self-Driving (Supervised).1 That split matters for reading the headline figure: the eight Cybercab crashes are not one fleet doing one kind of driving, but two very different kinds of operation counted together.
The four driverless crashes: all minor
Here is what Electrek found in the driverless file, all in Austin:1
- August: drove into a large water-filled pothole in a parking lot at 5 mph, with a safety monitor on board.
- September: stopped at a parking lot exit when a car backing out of a spot hit its right side.
- September: stopped in traffic and rear-ended by a pickup truck.
- September: drove over a raised concrete edge of a parking lot, across the sidewalk and down a curb at up to 3 mph, with a rider on board.
Electrek calls all four minor, says two weren’t Tesla’s fault at all, and notes that none involved an injury.1 In two of the four, the Cybercab was not moving when another vehicle struck it.1 For a brand-new vehicle carrying the public with nobody behind the wheel, that is a clean opening chapter.

The test fleet with a human at the wheel
The other half of the count comes from somewhere else entirely. Tesla has been testing Cybercabs fitted with steering wheels and a human behind them, often in places where it has no Robotaxi permit.1 Those crashes land in the ADAS file, with the driver type coded “In-Vehicle (Commercial / Test)” and the system “Verified Engaged.”1 There are four of them, all in September 2026:1
| City | Road | Speed | Hit | Injury |
|---|---|---|---|---|
| Philadelphia, PA | Highway | 53 mph | Passenger car | Unknown |
| Austin, TX | Highway | 5 mph | SUV | Minor, hospitalized |
| Lehi, UT | Intersection | 11 mph | SUV | Minor, not hospitalized |
| Tampa, FL | Highway | 0 mph (stopped) | Passenger car | Property damage |
Tesla redacted the narrative of all four as confidential business information, so the public data does not show who was at fault or what the system was doing.1 Pennsylvania State Police, Austin Police, Lehi Police and Florida Highway Patrol are listed as the investigating agencies.1
Electrek says the Philadelphia crash is not the Broad Street incident it reported in October; this one happened on a highway in September, at around 10 p.m. local time.1 Tesla has no permit to run automated vehicles in Pennsylvania and does not run Robotaxi in Utah, where a Cybercab was photographed in Park City in September with a driver visible behind the wheel.1 Two of the eight Cybercab crashes involved injuries, and both are in this Level 2 file.1
Tesla’s position: zero injuries in driverless miles
Tesla’s own record points the same way as the driverless file. The Robotaxi Safety Report Tesla published on October 21 claims zero people injured by Robotaxi over 3,112,541 driverless miles between January 22 and October 18.1 Electrek notes that the report does not mention the Cybercab’s test fleet and excludes every mile driven with a safety monitor on board.1
Across all 43 Tesla Robotaxi reports in the ADS file, the highest pre-crash speed is 29 mph: a driverless Model Y with two riders hitting a pothole in Dallas in September, after which one rider said they sought medical care.1 The median speed across all 43 is 1 mph, and not one happened on a highway; Tesla coded 21 on streets, 15 in parking lots and 7 at intersections.1 Electrek reads those numbers as a sign of where Tesla lets the system drive, since Robotaxi sticks to low-speed routes inside small geofenced zones and largely stays off highways.1
The low-severity pattern is not new. In May, fan account The Tesla Newswire summarized the NHTSA reports then on file as 17 total events, including 5 where another road user struck the Tesla, with no major crashes or serious injuries reported.2 A legal explainer from GB Law adds an important reading note: the federal reporting duty turns on the crash and its outcome, not on who caused it, and many reported Tesla crashes involved another driver striking the Tesla.3
Elon Musk did not address the filings directly in the posts available to ENN. On the same day, he wrote on X: “Robotaxi life is a hard life. Tesla team empathizes with this guy.”4 The post does not say what it refers to.
How it fits the Robotaxi rollout
The Cybercab arrives on top of a service that has grown in careful steps. Tesla launched Robotaxi in Austin on June 22, 2025, with a Tesla employee safety monitor in the passenger seat; limited rides with no monitor began January 22, 2026, and Tesla said unsupervised service covered the Austin metro on June 3, 2026.3 Under Texas law, when a robotaxi’s automated driving system is engaged, that system is treated as the vehicle’s operator.3
The first production Cybercab rolled off the line at Gigafactory Texas on February 17, 2026, a two-seater without a steering wheel or pedals, and Musk hailed the milestone as ahead of schedule.5 At that point, one investing writer was already counting 14 reported crashes across the young robotaxi fleet and asking whether they could slow the Cybercab down.5 Nine months later the vehicle is carrying riders, and its driverless crashes on file remain low-speed events.1

Electrek’s broader argument is about what happens when the speed cap comes off. Its master dataset of Tesla’s Level 2 reports holds 69 fatal crashes; among the 52 with a recorded speed, the median was 53 mph and 29 happened at 50 mph or more.1 ENN covered some of those filings in our Autopilot and FSD crash report. Electrek itself stresses that engagement within 30 seconds of a crash, which is what triggers these reports, does not prove the system caused it.1 It describes the Cybercab test fleet as the place where low-speed geofenced driving and higher-speed driving overlap, with the one Cybercab crash at real speed being the 53 mph Philadelphia highway case.1
What to watch next
- The next NHTSA release. The ADS file jumped from 22 Tesla reports in July to 43 this month; the next update will show whether the driverless Cybercab record stays this mild as rides continue.1
- The redacted narratives. With all four test-fleet narratives marked confidential, fault in the Philadelphia, Austin, Lehi and Tampa crashes cannot be judged from the public data.1
- Testing beyond the geofence. Tesla is testing Cybercabs with drivers in states where it has no Robotaxi service, which is where the faster crashes on file come from.1
- Tesla’s next safety report. The October 21 report’s zero-injury figure covers driverless miles only, so any future report that adds the test fleet would give a fuller picture.1
For now, the federal filings tell a fairly simple story: when the Cybercab drives itself with nobody at the wheel, its recorded crashes have been slow, minor and injury-free.1



