Why the October 6th TCMV vote is postponed 👇🏼
Reuters is running a Belgian NGO report: "Tesla FSD often misreads speed limits."
It went to regulators 4 weeks before the EU vote on October 6
The claim: revoke FSD's EU approval
I read all 36 pages and here's what the headlines leave out
The test:
1 car, 1 driver, who also wrote the report
41 hours, including a bank holiday, about 400 km
This “research” was not peer reviewed, only reviewed by advocacy groups
Tesla's European safety data covers 100+ million km. One of these is being used to argue for revoking an EU approval.
The headline number
"FSD averaged 44 km/h in 30 zones."
Read their own definitions. "Segment Speed" is the MAXIMUM speed reached in a segment. So 44 km/h is an average of peaks, not how fast the car drove through the zone.
Then those peaks get plugged into WHO crash-impact curves to claim "6x fatality risk."
That's not how that data works at ALL!
The "pattern"
Their own tables:
Flanders: the 3 segments over the limit averaged 33 km/h.
Wallonia: 0 of 4 segments over the limit.
This is not a Belgium problem. It's a Brussels map-data problem on roads that changed from 50 to 30. Real, fixable, not a system failure.
The missing baseline
No measurement of surrounding traffic. No comparison with human drivers in the same zones.
Meanwhile, their own report says FSD was "careful and courteous" with pedestrians and cyclists, yielding even when it didn't have to.
Careful around vulnerable road users, but also an "unacceptable risk"? Pick one.
Cycle streets: the one fair point
Belgium bans overtaking cyclists in cycle streets. The Netherlands, where FSD was approved, doesn't. FSD tried to overtake, and that needs fixing.
But every time, the driver saw it coming and stopped it. That's supervised FSD working exactly as designed.
The fix, not the ban
Every finding points to a software fix: trust the camera over outdated map data, and add Belgian cycle-street rules.
Real non-compliance goes through market surveillance by the authorities, not through an NGO PDF two weeks before a vote.
Even Belgium isn't panicking. The Flemish transport ministry calls it "a learning system" with the driver "100% responsible." The RDW has the report and is handling it through normal monitoring.
I want FSD held to a high bar. I test it across Europe myself.
But one car, 400 km, peak speeds sold as averages, and zero human baseline isn't that bar.
Fix it, validate it and move on. Not revoke it.