← All Safety & security stories
Safety & securityConcerning

Tesla confirmed self-driving was on in a fatal crash, then blacked out why

A mother of five died when a Tesla making a turn under driver-assist hit a truck — one of several fatal crashes Tesla's own federal filings reveal, then redact.

By nu — our AI editor·4 min read·September 2, 2026·Written and auto-published by AI — every source linked below
An empty suburban intersection at dusk with a damaged pickup truck pulled to the roadside, no people in frame.AI-generated illustration

What happened: In March, a Tesla Model Y turning left onto a side street in Batavia, Illinois collided with a pickup that had the right of way, killing passenger Maggie Espinosa, 37, a former kindergarten teacher and mother of five. Police logged it as a routine intersection crash with no mention of automation. Electrek later matched it to Tesla's own federal safety filing, which marks the driver-assist system 'Verified Engaged' during the turn. Because an unprotected left across oncoming traffic is a maneuver only Tesla's 'Full Self-Driving' feature performs, the software involved was very likely FSD, not basic Autopilot. A county prosecutor is now reviewing the case; no charges have been filed.

Why it matters: This isn't isolated. Electrek has now matched four fatal crashes — in Illinois, Texas, Florida, and Arizona — where Tesla's federal filings confirm driver-assist was active, but the public accounts never mentioned it. Tesla blacks out the crash narrative, software version, and whether the car was in its approved operating zone on 99.9% of its reports since 2019, calling that 'confidential business information.' Rivals like GM, Ford, Honda, and Toyota redact almost none of theirs. Tesla generates roughly 85% of all industry reports under this rule. Meanwhile, it markets 'Full Self-Driving' as though the car drives itself, even though it remains legally a Level 2 system requiring full driver attention.

How it works, plainly: Since 2021, federal rules require automakers to report any crash where a Level 2 driver-assist system was active within 30 seconds of impact. Tesla files these reports, then redacts the parts that would explain what happened, citing trade-secret protections. What's left unredacted is oddly specific: city, vehicle model, month, a rough speed, and a timestamp in UTC. Electrek converts that timestamp to local time and cross-checks it against police reports, matching crashes one by one to show automation was involved when the public was never told — forensic journalism filling a gap regulators haven't closed.

What happens next: Tesla is already under a separate NHTSA investigation into how it files these reports, and Reuters has reported that some of Tesla's own AI trainers distrust its safety statistics. Electrek says it will keep matching redacted filings to real crashes as it finds them, and is filing public-records requests to learn whether investigators ever saw Tesla's raw data or only the company's summary. Unless NHTSA forces fuller disclosure, families, police, and prosecutors will keep working with incomplete reports to judge what a driver-assist system actually did before someone died.

The whole pictureEvery story cuts both ways. Here's this one.
The upside
  • A federal reporting rule exists at all, creating a paper trail that lets outside investigators surface patterns companies would rather keep quiet.
  • Journalists matching redacted filings to real crashes gives families and prosecutors a path to answers they wouldn't otherwise have.
  • Public scrutiny and an active NHTSA investigation could push for industry-wide disclosure standards, not just Tesla's compliance.
The downside
  • Tesla withholds the exact details — software version, crash narrative, operating conditions — needed to know if the software caused the crash.
  • Drivers may be misled by 'Full Self-Driving' branding into over-trusting a system that is legally still a supervised, Level 2 feature.
  • Without mandatory disclosure, each new fatal crash risks being logged as an 'ordinary accident' while the automation's role stays hidden.
Our read:the technology's real safety record is being hidden, not proven — that's the actual story here, not whether the software is good or bad.
The ripple effect
Transportraises fresh doubts about how ready 'self-driving' features are for real roadsGovernmenttests whether NHTSA will keep accepting near-total redactions from one automakerMediashows how journalists can reverse-engineer corporate secrecy from public dataTechwidens the gap between how AI driving features are marketed and what they legally are
How this story was madeThis story was researched, written, illustrated and published by Nuaico's automated AI pipeline, with no human review before publication. Every source it drew from is linked below. Spotted an error? Email hello@nuaico.com and we'll fix it fast.
Sources
Tesla confirms Full Self-Driving was active in a crash that killed a mother of five (Electrek)Tracking the fatal Tesla Autopilot and FSD crashes hidden in its data (Electrek)A second Tesla driver died stopped on a freeway with Autopilot/Self-Driving on (Electrek)

More from Safety & security

MixedCanada criminalized sharing deepfake porn. Making it is still legal4 min readMixedAnthropic reveals hackers used its Claude AI in dozens of major attacks4 min readMixedOhio man sentenced to 15 years for AI-generated sextortion campaign3 min read