Tesla presented misleading Full Self-Driving crash data to European regulators, Reuters finds

The data in question

Tesla claims its Full Self-Driving system is up to ten times safer than a human driver. In a slide deck shown to European authorities, the company said cars running FSD travel more than seven times farther between crashes than the average U.S. driver. One presentation even suggested that replacing every American vehicle with an FSD-equipped Tesla would prevent 32,000 traffic deaths a year.

We are certainly concerned that Tesla presented unreliable safety data from the United States to regulators in Sweden.

Researchers told Reuters the figures rest on invalid comparisons, counting only crashes severe enough to deploy an airbag for Tesla, while using broad national figures that include minor accidents.

Approval in the Netherlands

Tesla first approached the Dutch road regulator, RDW, in late 2024. A November 2024 letter pointed to the company’s own safety report and said increased FSD usage “leads to safer roads.” After more than a year of testing, RDW approved FSD for use in the Netherlands in April 2026, making it the first EU country to do so. The regulator is now seeking EU-wide approval on Tesla’s behalf.

1. Sep 1, 2024Tesla approaches RDW to begin FSD approval process
2. Nov 1, 2024Tesla sends letter to RDW with safety report, claims safer roads
3. Apr 10, 2026RDW approves FSD for use in the Netherlands
4. Apr 15, 2026Tesla policy manager contacts Swedish regulators with slide deck
5. Jun 15, 2026Reuters publishes investigation revealing misleading statistics

Swedish push and watchdog concern

Soon after the Dutch decision, Tesla policy manager Ivan Komusanac wrote to Swedish regulators asking for similar approval and attached the safety claims. The European Transport Safety Council said the numbers were presented inside a regulatory process where officials must weigh whether the system is safe enough to deploy more widely.

What the comparisons missed

Researchers identified two main distortions. First, Tesla’s own figures count only crashes where an airbag went off, while the U.S. crash data it compares against includes every reported incident, from fender bumps upward. Second, a Tesla with FSD is a new car, but the U.S. fleet averages 13 years old, newer cars already have safety features that lower crash rates, making the like-for-like comparison flattering.

Tesla FSD (company claim)

0.182

U.S. average all vehicles

1.515

Tesla FSD (company claim)

0.182 crashes per million miles

U.S. average all vehicles

1.515 crashes per million miles

The hypothetical 32,000-lives figure simply assumes every U.S. vehicle is replaced by an FSD Tesla that is seven times safer, a calculation one expert called “less a measurement than a hypothetical.”

Broader FSD doubts

European officials had already raised concerns about FSD’s behaviour on icy roads, its tendency to speed, and driver distraction. Dutch regulators have questioned whether the name “Full Self-Driving” overpromises for a system that still requires a human to supervise. Meanwhile, some Tesla owners in the Netherlands complain the feature does not work on older vehicles. RDW says it does not rely on marketing claims or external statistics and performs its own tests, analyses and verifications. Tesla did not respond to requests for comment.