US auto safety regulators are investigating incidents involving Tesla’s self-driving robotaxis during their first day of paid operations in Austin, Texas.

The National Highway Traffic Safety Administration (NHTSA) confirmed it is gathering information from Tesla after social media videos appeared to show the vehicles violating traffic laws.

One video, posted by Tesla investor Rob Maurer, depicted a Model Y entering an intersection from a left-turn-only lane.

The vehicle hesitated, then swerved into oncoming traffic before returning to the correct lane over a double-yellow line.

Other videos showed Tesla robotaxis exceeding speed limits, including a vehicle traveling at 35 miles per hour in a 30 mph zone.

In a livestream, a Model Y was observed driving at 39 mph in a 35 mph zone, matching the speed of surrounding traffic. Another video showed a driverless Model Y stopping in the middle of the road after a passenger attempted to activate a feature to pull over.

The NHTSA said it will evaluate these reports and take necessary actions to ensure road safety.

The agency is also investigating Tesla’s full self-driving system for performance issues under conditions with limited visibility, such as glare or fog.

Neither Tesla nor the Austin Police Department has commented on the incidents.

.source-ref{font-size:0.85em;color:#666;display:block;margin-top:1em;}a.ask-tia-citation-link:hover{color:#11628d !important;background:#e9f6f5 !important;border-color:#11628d !important;text-decoration:none !important;}@media only screen and (min-width:768px){a.ask-tia-citation-link{font-size:11px !important;}}

🔗 Source: Bloomberg

🧠 Food for thought

1️⃣ Tesla faces intensified scrutiny amid broader autonomous vehicle safety record

The recent incidents in Austin represent just one chapter in Tesla’s complex history with regulators. Tesla vehicles have consistently dominated autonomous vehicle incident reports, accounting for 53.9% of all reported autonomous vehicle accidents between 2019 and 2024 according to NHTSA data 1.

This pattern has led to regulatory action before. Tesla recalled over 362,000 vehicles in February 2023 because NHTSA determined its driver-assistance system might allow cars to violate traffic laws, mirroring the exact issues observed in Austin 1.

The Austin incidents reflect a fundamental challenge for autonomous systems: navigating complex traffic scenarios. Research shows that while autonomous vehicles generally have lower accident rates, they struggle specifically with turning at intersections and dawn/dusk conditions—precisely the type of situation where the Model Y in Austin swerved into an oncoming lane 2.

This creates a regulatory dilemma for NHTSA, which must balance allowing technological advancement while ensuring public safety, particularly as Tesla’s approach differs from competitors who typically test with safety drivers or in more controlled environments.

2️⃣ Reporting requirements create perception challenges for autonomous vehicle safety

The incidents in Austin represent a common challenge in evaluating autonomous vehicle safety: isolated incidents receive outsized attention while broader safety data tells a more complex story.

Comprehensive safety analysis of autonomous vehicles shows mixed results that aren’t reflected in individual incident reporting. The Nature Communications study analyzing 2,100 autonomous vehicles found they had significantly lower rates of accidents caused by inattention or poor driving (1.8% for autonomous vehicles versus 19.8% for human-driven vehicles) 2.

However, California data shows autonomous vehicles recorded 96.7 crashes per 1,000 vehicles in 2022, compared to just 7.0 for conventional vehicles—a statistic that appears concerning until accounting for stricter reporting requirements for autonomous vehicles, which must report all collisions regardless of severity 3.

This discrepancy creates a perception challenge as companies like Tesla face intense public scrutiny for every incident while the broader safety benefits may only become apparent through long-term data analysis, not viral videos of individual failures.

……