This whole chain is a response to

Teslas are programmed to turn off autopilot half a second before a crash so statistics can say “autopilot was off at the time of the crash”.

And no one has anything to support that.

Please add a citation to support that before you insist on citations to support disagreeing with an unsupported claim

Note my previous response (accidentally responded to the nhtsa thing in the past with an edit)

Im not disagreeing that FSD can disengage just before a crash, just arguing that I’ve not seen them try to shift blame based on it and I remember reading somewhere that it’s intentional to try to mitigate the aftermath