eyes on the road

US to probe Texas-based Tesla's self-driving system after pedestrian killed in low visibility conditions

Investigators will look into the ability of “Full Self-Driving” to “detect and respond appropriately to reduced roadway visibility conditions, and if so, the contributing circumstances for these crashes." Photo courtesy of Tesla

The U.S. government's road safety agency is investigating Tesla's “Full Self-Driving” system after getting reports of crashes in low-visibility conditions, including one that killed a pedestrian.

The National Highway Traffic Safety Administration said in documents that it opened the probe last week after the company reported four crashes when Teslas encountered sun glare, fog and airborne dust.

In addition to the pedestrian's death, another crash involved an injury, the agency said.

Investigators will look into the ability of “Full Self-Driving” to “detect and respond appropriately to reduced roadway visibility conditions, and if so, the contributing circumstances for these crashes.”

The investigation covers roughly 2.4 million Teslas from the 2016 through 2024 model years.

A message was left Friday seeking comment from Tesla, which has repeatedly said the system cannot drive itself and human drivers must be ready to intervene at all times.

Last week Tesla held an event at a Hollywood studio to unveil a fully autonomous robotaxi without a steering wheel or pedals. Musk, who has promised autonomous vehicles before, said the company plans to have autonomous Models Y and 3 running without human drivers next year. Robotaxis without steering wheels would be available in 2026 starting in California and Texas, he said.

The investigation's impact on Tesla's self-driving ambitions isn't clear. NHTSA would have to approve any robotaxi without pedals or a steering wheel, and it's unlikely that would happen while the investigation is in progress. But if the company tries to deploy autonomous vehicles in its existing models, that likely would fall to state regulations. There are no federal regulations specifically focused on autonomous vehicles, although they must meet broader safety rules.

NHTSA also said it would look into whether any other similar crashes involving “Full Self-Driving” have happened in low visibility conditions, and it will seek information from the company on whether any updates affected the system’s performance in those conditions.

“In particular, this review will assess the timing, purpose and capabilities of any such updates, as well as Tesla’s assessment of their safety impact,” the documents said.

Tesla reported the four crashes to NHTSA under an order from the agency covering all automakers. An agency database says the pedestrian was killed in Rimrock, Arizona, in November of 2023 after being hit by a 2021 Tesla Model Y. Rimrock is about 100 miles (161 kilometers) north of Phoenix.

The Arizona Department of Public Safety said in a statement that the crash happened just after 5 p.m. Nov. 27 on Interstate 17. Two vehicles collided on the freeway, blocking the left lane. A Toyota 4Runner stopped, and two people got out to help with traffic control. A red Tesla Model Y then hit the 4Runner and one of the people who exited from it. A 71-year-old woman from Mesa, Arizona, was pronounced dead at the scene.

The collision happened because the sun was in the Tesla driver's eyes, so the Tesla driver was not charged, said Raul Garcia, public information officer for the department. Sun glare also was a contributing factor in the first collision, he added.

Tesla has twice recalled “Full Self-Driving” under pressure from NHTSA, which in July sought information from law enforcement and the company after a Tesla using the system struck and killed a motorcyclist near Seattle.

The recalls were issued because the system was programmed to run stop signs at slow speeds and because the system disobeyed other traffic laws. Both problems were to be fixed with online software updates.

Critics have said that Tesla’s system, which uses only cameras to spot hazards, doesn’t have proper sensors to be fully self driving. Nearly all other companies working on autonomous vehicles use radar and laser sensors in addition to cameras to see better in the dark or poor visibility conditions.

Musk has said that humans drive with only eyesight, so cars should be able to drive with just cameras. He has called lidar (light detection and ranging), which uses lasers to detect objects, a “fool's errand.”

The “Full Self-Driving” recalls arrived after a three-year investigation into Tesla's less-sophisticated Autopilot system crashing into emergency and other vehicles parked on highways, many with warning lights flashing.

That investigation was closed last April after the agency pressured Tesla into recalling its vehicles to bolster a weak system that made sure drivers are paying attention. A few weeks after the recall, NHTSA began investigating whether the recall was working.

NHTSA began its Autopilot crash investigation in 2021, after receiving 11 reports that Teslas that were using Autopilot struck parked emergency vehicles. In documents explaining why the investigation was ended, NHTSA said it ultimately found 467 crashes involving Autopilot resulting in 54 injuries and 14 deaths. Autopilot is a fancy version of cruise control, while “Full Self-Driving” has been billed by Musk as capable of driving without human intervention.

The investigation that was opened Thursday enters new territory for NHTSA, which previously had viewed Tesla's systems as assisting drivers rather than driving themselves. With the new probe, the agency is focusing on the capabilities of “Full Self-Driving" rather than simply making sure drivers are paying attention.

Michael Brooks, executive director of the nonprofit Center for Auto Safety, said the previous investigation of Autopilot didn't look at why the Teslas weren't seeing and stopping for emergency vehicles.

“Before they were kind of putting the onus on the driver rather than the car,” he said. “Here they're saying these systems are not capable of appropriately detecting safety hazards whether the drivers are paying attention or not.”

Trending News

A View From HETI

Chevron is in talks with Microsoft and Engine No. 1 about a massive natural gas power plant in Texas. Photo via Getty Images

Software giant Microsoft is negotiating exclusively with Houston-based oil and gas titan Chevron and investment firm Engine No. 1 about the development of a $7 billion power plant in West Texas that would supply electricity for a Microsoft data center campus.

The proposed natural-gas-fired plant initially would generate 2,500 megawatts of electricity, Bloomberg reports. The plant would be built near Pecos, a Permian Basin city, in an area where Microsoft plans to build a 2,500-megawatt data center campus on a 7,000-acre site.

A deal with Microsoft would secure a long-term customer for the plant’s output and help finance its construction, Bloomberg says. The project, expected to be producing power by 2030, still requires tax and environmental approvals as well an agreement to terms among Chevron, Engine No. 1, and Microsoft.

In a statement issued after Bloomberg reported the news, Chevron acknowledged it was in exclusive talks with Engine No. 1 and Microsoft, but the oil and gas company offered no details.

Chevron says the proposed plant “reflects an emerging shift in how power for AI is being developed, bringing energy supply closer to demand through co-located, behind-the-meter generation to deliver reliability while helping avoid added strain on regional electricity systems. It pairs sustained, always-on demand from advanced computing with proven capability to design, build, and operate large-scale energy infrastructure.”

Development of gas-powered electrical plants for AI data centers represents a new—and potentially lucrative— business line for Chevron. In 2025, Chevron, Engine No. 1 and GE Vernova announced a partnership to produce natural gas for AI data centers in the U.S.

Chevron’s collaboration with Engine No. 1 has already secured an order for seven large natural gas turbines from GE Vernova, according to Bloomberg.

“Energy is the key to America’s AI dominance,” Chris James, founder and chief investment officer of Engine No. 1, said last year. “By using abundant domestic natural gas to generate electricity directly connected to data centers, we can secure AI leadership, drive productivity gains across our economy, and restore America’s standing as an industrial superpower.”

Trending News