I am mainly talking about the first 5-10 years of autonomous cars being available. Beyond that, I won't even attempt to make any predictions. During this first

I am mainly talking about the first 5-10 years of autonomous cars being available. Beyond that, I won't even attempt to make any predictions. During this first decade, most cars on the roads will still be human-driven. I don't see the insurance end-user cost being decreased much for the autonomy-capable car owners. If anything, I see it costing more during the very first years until there's more statistics & operational experience. So even if there's some price reduction during the later years of the first decade, I think it will probably only even out with the initial premium, at best. As for new overhead costs, I guess during the initial years there could be a learning period of how to handle crash drama where the car owners don't agree as easily to what happened and who's at fault. Those same first years the A.I. will probably be pretty buggy as well, with some surprising crashes. However I think the insurance companies will account for these new overheads and raise the price accordingly. Overall, I don't think investing in car insurance is a goldmine opportunity. I do think that they will continue to make nice profits and definitely shouldn't be sold short. At least not for the reason of autonomous cars by themselves, maybe sell short a selection of companies whose management you think will be unable to adapt.