This is just such a bad take, and it's so disappointing to see it parroted all over the web. So many things are just completely inaccurate about these "statistics", and it's probably why it "seems" so many are against autonomous vehicles.

1. These are self-reported statistics coming from the very company(s) that have extremely vested interests in making themselves look good.
2. These statistics are for vehicles that are currently being used in an extremely small (and geo-fenced) location(s) picked for their ability to be the easiest to navigate while being able to say "hey we totally work in a big city with lots of people".

• These cars don't even go onto highways or areas where accidents are more likely.
• These cars drive so defensively they literally shut down so as to avoid causing any accidents (hey, who cares if we block traffic and cause jams because we get to juice our numbers).

3. They always use total human driven miles which are a complete oranges to apples comparison: Their miles aren't being driven

• In bad weather
• On dangerous, windy, old, unpaved, or otherwise poor road conditions
• In rural areas where there are deer/etc that wander into the road and cause accidents

4. They also don't adjust or take any median numbers as I'm not interested in them driving better than the "average" driver when that includes DUIs, crashes caused by neglect or improper maintenance, reckless drivers, elderly drivers, or the fast and furious types crashing their vehicle on some hill climb driving course.
5. And that's all just off the top of my head.

So no, I would absolutely not say they are "less prone to accidents than human drivers". And that's just the statistics, to say nothing about the legality that will come up. Especially given just how adverse companies seem to be to admit fault for anything.