Submit incident
Documented

One Number From NHTSA Should Worry Every Driver-Assist Buyer

June 15, 2022
Curated by Team Raidu · Reviewed by Shiva Ganesh
oecd:2022-06-15-1c5bView source ↗
LinkedInX

What happened

The National Highway Traffic Safety Administration is not a body prone to alarm. So when the agency's own crash figures show one automaker responsible for roughly seven out of every ten incidents tied to automated driving features, that is not noise, it is a signal. Between July 2021 and May 2022, Tesla vehicles dominated the count of US crashes involving driver-assist systems, and they dominated the count of the worst outcomes too: the injuries that put people in hospitals and the ones that killed them.

Scale explains part of this. Tesla has more vehicles on the road running these features than any rival, and a bigger fleet will naturally generate more incidents. But scale alone does not account for a company pulling nearly seventy percent of a category's crash total, nor does it explain the skew toward severe injury and fatal outcomes rather than minor fender-benders. If the system were performing in line with its market share, the injury and fatality share should track closer to that share, not run ahead of it.

NHTSA's own caveat is the more interesting part of the story. The agency said plainly that the data lacks the context needed to draw firm conclusions, an unusually candid admission from a regulator that just published numbers implicating a specific company. Read charitably, it means reporting requirements, crash narratives, and telemetry standards have not caught up to what these systems actually do on the road. Read less charitably, it means nobody, not the regulator, not the public, has a reliable way to tell whether the driver was engaged, whether the software made the call, or which one deserves blame when the outcome is fatal.

That gap is the real finding here. A safety statistic without a verifiable chain back to system behavior cannot settle the question it raises, it can only gesture at it. Until crash data captures what the AI actually did in the moments before impact, and who had the chance to intervene, headline numbers like this one will keep drawing scrutiny without resolving it.

Reported impact

Affected parties
Not publicly disclosed
Harm type
Not publicly disclosed
Scale
Not publicly disclosed
Financial impact
Not publicly disclosed
Regulatory action
Not publicly disclosed

Classification

Organization
Not publicly disclosed
AI system
Not publicly disclosed
Industry
Not publicly disclosed
Country
Not publicly disclosed
Provider
Not publicly disclosed
Incident type
Not publicly disclosed

Relevant governance controls

Governance control mapping is not available for this record.

  • No controls mappedNot publicly disclosed

Control mapping is analytical. It does not state that any control would have prevented the incident.

Sources and evidence

This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.

OECD AI Incidents Monitor
Also catalogued in
One Number From NHTSA Should Worry Every Driver-Assist Buyer
2022-06-15