The first Tesla robotaxi crash in Austin happened just weeks after the service launched. The most recent batch of five happened in a single month. In between, the pattern stayed consistent: federal data shows Tesla’s autonomous fleet is hitting things at a rate that would get a teenage driver’s insurance canceled.

Bloomberg’s Kara Carlson reported today that Tesla has filed 14 crash incidents with the National Highway Traffic Safety Administration since its Austin robotaxi service began in June 2025. The reports, filed under NHTSA’s Standing General Order for automated driving systems, paint a picture that clashes directly with CEO Elon Musk‘s January claim at Davos that self-driving is “essentially a solved problem.”

The Fact: Tesla’s Austin robotaxi fleet has been involved in 14 crash incidents across an estimated 800,000 miles of operation, or approximately one crash every 57,000 miles. Human drivers average one police-reported crash about every 500,000 miles, according to NHTSA.
The Delta: Tesla quietly updated a July 2025 crash report five months later to add that the incident caused minor injuries and hospitalization. That information was absent from the original filing. Every Tesla crash narrative in the NHTSA database is fully redacted as “confidential business information.” Waymo, Zoox, Aurora, and Nuro all provide detailed descriptions of their incidents.
The Buyer Impact: Tesla began offering driverless rides without any safety monitor in January 2026. That same month, the company reported four new crashes. Prospective robotaxi riders in Austin have no way to independently evaluate whether Tesla’s system was at fault in any of these incidents.

Five New Crashes in One Month Bring the Total to 14

Tesla submitted five new crash reports in January 2026, covering incidents from December 2025 and January 2026. All five involved Model Y vehicles operating with the autonomous driving system verified as engaged in Austin. The incidents included a collision with a fixed object at 17 mph while the vehicle was driving straight, a crash with an Austin city bus while the Tesla was stationary, a collision with a heavy truck at 4 mph, and two separate low-speed incidents where the robotaxi backed into objects in parking lots at 1 and 2 mph.

The January reports also revealed a quiet revision. A July 2025 crash that Tesla originally filed as property damage only was updated in December to include a minor injury requiring hospitalization. A second July incident also caused minor injuries. Bloomberg reported that information on all of Tesla’s incidents is limited, and unlike competitors, Tesla redacts the narrative descriptions that would explain what actually happened.

Some of these crashes may not be Tesla’s fault. A stationary robotaxi struck by a city bus, for example, is almost certainly the bus driver’s error. Low-speed backing incidents at 1-2 mph could be trivial. But without the narrative details that every other ADS operator provides to NHTSA, there is no way to tell which crashes resulted from system failures and which were caused by other road users.

The Crash Rate in Context: Tesla vs. Human Drivers vs. Waymo

Tesla’s robotaxi fleet has traveled an estimated 800,000 cumulative miles since the Austin service launched in late June 2025. That mileage estimate, based on extrapolation from Tesla’s Q4 2025 earnings data showing roughly 700,000 cumulative paid miles through November, was first calculated by Electrek’s Fred Lambert. With 14 reported crashes, that works out to one incident roughly every 57,000 miles.

How bad is that? It depends on the benchmark. NHTSA data shows the average American driver reports a police-recorded crash about once every 500,000 miles, which would make Tesla’s rate roughly nine times worse. But that figure only counts police-reported incidents. Including unreported fender-benders, estimates put the human average closer to one crash every 200,000 miles, making Tesla’s rate about 3.5 times worse. Tesla’s own Vehicle Safety Report cites a figure of one minor collision every 229,000 miles for typical drivers, putting its robotaxi fleet at roughly four times the rate Tesla itself considers normal.

The comparison with Waymo is harder to make apples-to-apples, but still instructive. Alphabet‘s autonomous vehicle unit operates about 200 vehicles in Austin through the Uber app and has reported approximately 50 incidents in the city, according to Bloomberg. Nationally, Waymo’s fleet of roughly 2,500 vehicles has logged over 200 million fully driverless miles as of early 2026, with no safety monitor, no chase car, and no human backup. Tesla’s Austin fleet, until late January, had a trained safety monitor in the passenger seat for nearly every mile driven. Crashes still occurred at a higher rate.

Austin Is Still Tesla’s Only Real Robotaxi Market

Eight months after launch, Austin remains the only city where Tesla offers anything resembling an autonomous ride-hailing service. The company launched a service in the San Francisco Bay Area on the same app, but those trips use human drivers because Tesla has not applied for the California DMV permits required for driverless testing or deployment. Bloomberg described the Bay Area service as “more akin to Uber.”

Tesla has not disclosed the current size of its Austin fleet. At the Q4 2025 earnings call in late January, Musk said the company had about 500 rideshare vehicles between Austin and the Bay Area combined. Sherwood News, citing the NHTSA data, reported the Austin fleet at approximately 45 vehicles. That is a long way from the 1,000 vehicles in Austin that Musk promised “within a few months” of launch.

Tesla says it aims to expand to seven more cities by mid-2026, including Dallas, Houston, Phoenix, Miami, Orlando, Tampa, and Las Vegas. But the company failed to meet similar expansion targets for the end of 2025. In July, Musk predicted the service would reach “half the population of the U.S.” by year-end. It didn’t come close. Waymo, which began offering driverless rides to the public in the Phoenix area in October 2020, already operates across six cities and plans to expand to at least 10 more.

The Transparency Gap Is as Troubling as the Crash Rate

Every other company in the NHTSA Standing General Order database provides detailed narrative descriptions of what happened in each crash. Waymo’s reports, for instance, typically explain the vehicle’s speed, direction, what the other party was doing, and who initiated contact. These accounts allow independent researchers and regulators to assess fault.

Tesla redacts everything. Every crash narrative in its filings is replaced with the same phrase: “REDACTED, MAY CONTAIN CONFIDENTIAL BUSINESS INFORMATION.” This matters because some of these 14 incidents were clearly not the robotaxi’s fault. Others might have been entirely preventable. Without narrative data, the public cannot distinguish between the two categories. Tesla is asking riders to trust its safety record while making independent verification impossible.

This matters more now than it did six months ago. After tests in December, Tesla began offering driverless rides without any safety monitor in “a few” vehicles in its Austin fleet in January. Bloomberg reported that it was not immediately clear whether any of the five January incidents involved vehicles operating without monitors.

EVXL’s Take

I’ve been covering Tesla’s Austin robotaxi program since before the first Model Y rolled out of Giga Texas in June 2025. The early driving errors we documented within days of launch, wrong-lane incursions, phantom braking, mid-intersection drop-offs, weren’t growing pains. They were early data points in what has become a consistent pattern.

The crash rate is bad. The redaction is worse. Waymo publishes full narratives for every incident. Tesla hides behind boilerplate confidentiality claims. When your fleet crashes at three to four times the human rate and you won’t show anyone what happened, the silence speaks louder than the data.

And here’s the timing that should concern everyone: Musk declared FSD “pretty much solved” in December, removed safety monitors from some vehicles in January, and in that same month Tesla filed four new crash reports. Three of those crashes involved backing into objects or colliding with stationary obstacles, the kind of low-speed errors that suggest the system still struggles with basic spatial awareness.

My prediction: Tesla will not meet its seven-city expansion target by mid-2026. New Texas robotaxi regulations require an Autonomous Vehicle Operation Permit from the state DMV before paid driverless rides can legally operate. According to analysis by Brad Munchen, the DMV’s new permitting system won’t be operational until May 28, limiting Tesla’s ability to even begin the application process. Meanwhile, Tesla still hasn’t applied for driverless testing permits in California. Expect the “expansion” story to shift quietly from “seven cities by mid-year” to “refining operations in existing markets” by the Q2 earnings call.

Musk has promised driverless Teslas every year since 2016. The pattern holds. What’s different now is that the data is public, the crash rate is quantifiable, and Waymo is operating at a completely different scale without hiding its homework.

Frequently Asked Questions

How many crashes has the Tesla robotaxi fleet had in Austin?

Tesla has reported 14 crash incidents to the NHTSA since the Austin robotaxi service launched in June 2025. Five of those were reported in January 2026 alone, covering incidents from December 2025 and January 2026.

How does Tesla’s robotaxi crash rate compare to human drivers?

Tesla’s fleet crashes approximately once every 57,000 miles. The average American driver reports a police-recorded crash about once every 500,000 miles according to NHTSA. Including unreported incidents, the estimate is closer to once every 200,000 miles. By the most conservative comparison, Tesla’s rate is roughly 3.5 times worse than human drivers.

Are Tesla robotaxis operating without safety monitors?

Yes. Tesla began offering driverless rides without a human safety monitor in a small number of Austin vehicles in January 2026. Musk announced this on X on January 22. Most vehicles in the fleet still operate with monitors present. Video evidence reported by Electrek suggests some “unsupervised” vehicles are followed by trailing cars with safety personnel inside.

When will Tesla expand robotaxi service to other cities?

Tesla has announced plans to expand to Dallas, Houston, Phoenix, Miami, Orlando, Tampa, and Las Vegas by mid-2026. However, the company missed similar expansion targets for the end of 2025 and faces new regulatory requirements in Texas that take effect in late May 2026.

Editorial Note: AI tools were used to assist with research and archive retrieval for this article. All reporting, analysis, and editorial perspectives are by Haye Kesteloo.