Back to stories
Policy

NHTSA Escalates Tesla FSD Investigation to Engineering Analysis Covering 2.4 Million Vehicles

Michael Ouroumis2 min read
NHTSA Escalates Tesla FSD Investigation to Engineering Analysis Covering 2.4 Million Vehicles

The National Highway Traffic Safety Administration has escalated its investigation into Tesla's Full Self-Driving software to a formal Engineering Analysis — the most serious level of federal auto safety probe — covering approximately 2.4 million vehicles.

What Triggered the Upgrade

The new Engineering Analysis (EA26002) upgrades a preliminary evaluation (PE24031) that NHTSA originally launched in October 2024 after identifying four crashes in reduced-visibility conditions, including one that fatally struck a pedestrian. The scope has since expanded to nine total incidents involving one fatality and one injury.

NHTSA investigators found that FSD's camera degradation detection system — the software designed to recognize when sensors cannot see properly and alert the driver — fails under common roadway conditions. According to the agency, the system "did not detect common roadway conditions that impaired camera visibility and/or provide alerts when camera performance had deteriorated until immediately before the crash occurred."

A Growing Pattern of Violations

The escalation comes amid a broader pattern of FSD-related safety concerns. NHTSA has identified 80 Tesla FSD traffic violations to date, up from 50 in October, spanning incidents that include running red lights and crossing into opposing lanes. These findings are drawn from 62 driver complaints, 14 Tesla-filed reports, and 4 media accounts.

Tesla has also faced scrutiny over delays in providing FSD traffic violation data to the agency. Earlier this year, NHTSA granted the company a five-week extension to review over 8,000 potential traffic violations.

What an Engineering Analysis Means

NHTSA typically completes an Engineering Analysis within 18 months. The process involves deeper technical testing, additional information requests, and potential real-world evaluations. Historically, an EA is the final investigative step before the agency either closes a case or pushes for a recall.

For Tesla, the stakes are significant. A recall order could force a software update or even a suspension of FSD capabilities across millions of vehicles, directly impacting the company's autonomous driving ambitions.

Broader Implications

The probe intensifies at a pivotal moment for Tesla's self-driving program. CEO Elon Musk recently announced plans for robotaxi service expansion and confirmed that the company is nearing a tapeout of its next-generation AI6 chip. However, regulatory pressure from NHTSA could complicate the timeline for broader autonomous deployment.

The investigation also raises wider questions for the autonomous vehicle industry about how camera-based systems handle adverse weather and visibility conditions — a challenge that competitors using lidar-based approaches have long cited as a limitation of vision-only architectures.

Learn AI for Free — FreeAcademy.ai

Take "AI Essentials: Understanding AI in 2026" — a free course with certificate to master the skills behind this story.

More in Policy

Chinese Court Rules Firing Workers to Replace Them With AI Is Illegal
Policy

Chinese Court Rules Firing Workers to Replace Them With AI Is Illegal

The Hangzhou Intermediate People's Court has ruled that companies cannot lawfully dismiss employees solely to replace them with artificial intelligence, classifying AI adoption as a strategic business choice rather than an unavoidable change in circumstances.

17 hours ago3 min read
NIST CAISI: China's DeepSeek V4 Pro Trails US Frontier AI by Eight Months
Policy

NIST CAISI: China's DeepSeek V4 Pro Trails US Frontier AI by Eight Months

NIST's Center for AI Standards and Innovation finds DeepSeek's most capable model performs on par with GPT-5 from eight months ago, but at a meaningful cost advantage on most benchmarks.

20 hours ago2 min read
Academy Rules AI-Generated Actors and Screenplays Ineligible for Oscars
Policy

Academy Rules AI-Generated Actors and Screenplays Ineligible for Oscars

The Academy of Motion Picture Arts and Sciences announced new eligibility rules barring AI-generated performances and screenplays from competing for Oscars, with acting roles required to be 'demonstrably performed by humans with their consent.'

1 day ago2 min read