Tesla robotaxi ambitions face safety reality

Tesla’s robotaxi programme is facing renewed scrutiny after crash data filed with the National Highway Traffic Safety Administration showed that its self-driving ride-hailing service remains a limited and closely monitored operation, despite repeated claims that autonomy would soon become a central pillar of the company’s growth.

The disclosures show 17 reported incidents involving Tesla robotaxis between July 2025 and March 2026, most of them low-speed crashes in Austin, Texas. Two of the incidents involved vehicles being controlled remotely by Tesla employees after onboard safety monitors sought assistance, highlighting the continuing role of human oversight in a service marketed as a major step towards driverless transport.

One crash occurred in July 2025 after a robotaxi stopped by the side of a street and failed to move forward. A remote operator took control and drove the vehicle up a kerb and into a metal fence at about 8mph. The safety monitor in the passenger seat suffered minor injuries but was not hospitalised. Another incident in January 2026 involved a remote driver taking over after navigation assistance was requested, before the vehicle struck a temporary construction barricade at about 9mph.

The details are significant because Tesla’s strategy differs from several other autonomous-vehicle developers. While most operators use remote support teams to provide guidance to a vehicle’s software, Tesla’s filings indicate that its remote staff have at times directly driven robotaxis out of difficult situations. That practice has drawn attention to questions over visibility, network latency and the situational awareness available to an operator who is not physically inside the vehicle.

Tesla’s robotaxi service is currently operating in Austin, Dallas and Houston, but the scale remains modest. The company is estimated to have fewer than 100 robotaxi vehicles across those markets, with some still using safety monitors. In Austin, the fleet is understood to be around 50 vehicles, compared with more than 250 Waymo vehicles operating in the same city. The gap underlines how far Tesla remains from the large-scale deployment envisaged by Elon Musk, who has repeatedly argued that self-driving capability could transform the company’s valuation and business model.

The operational record also points to a service still finding its footing. Riders in Dallas and Houston have encountered long wait times, limited vehicle availability and drop-off points some distance from intended destinations. A five-mile ride in Dallas reportedly took far longer than a conventional trip because of repeated attempts to secure a car, routing choices that avoided a major expressway and a final drop-off requiring a walk to the destination. In Houston, the operating zone remains narrow, with ride requests sometimes cancelled after a vehicle appears to be available.

Tesla has framed the measured rollout as a cautious approach intended to avoid serious injuries or fatalities. That argument carries weight in a sector where public confidence can be damaged by a single high-profile crash. The disclosed incidents do not show a pattern of severe collisions, and several appear to involve minor contact or property damage. Austin police officials have not reported major Tesla robotaxi crashes or traffic citations linked to the programme.

Yet the filings weaken the perception that Tesla is close to an effortless nationwide robotaxi launch. Federal crash-reporting rules require companies to disclose certain incidents involving automated driving systems, including crashes where the system was engaged shortly before impact and where specific damage or injury thresholds are met. The data does not, by itself, provide a full crash-rate comparison because it does not include total miles travelled, fleet size, exposure by road type or comparable operating conditions.

Tesla has argued that raw crash counts can be misleading because connected vehicles may report more incidents than less connected fleets. The company says its telemetry allows it to detect and report events that other manufacturers might not capture unless informed by drivers, customers or legal claims. That context is important, but it does not remove the central issue facing the robotaxi programme: the business still depends on human backstops, restricted service zones and careful operating conditions.

Competition is intensifying as Waymo expands its own robotaxi network, while also dealing with regulatory and safety challenges of its own. The broader industry remains under pressure to prove that autonomous vehicles can handle unusual road layouts, construction zones, severe weather, vulnerable road users and unpredictable human drivers.