There is no reliable citation that you can give to prove that. Emphasis on reliable, as Tesla's own stats are highly misleading -- they basically compare highw

There is no reliable citation that you can give to prove that. Emphasis on reliable, as Tesla's own stats are highly misleading -- they basically compare highway data, because Autopilot is mostly used on highways, to all other conditions in which accidents happen. I'm sure that if Autopilot was active all the time, the rate of accidents would be far higher, and Tesla's numbers wouldn't look so positive.