Sure the passing driver is ultimately at fault, but the slow driver has blood on their hands despite no accountability. If they were driving the legal speed limit it may have prevented a serious accident and loss of life. of course, the insurance company would be none the wiser.
If the slow driver is on a highway with a minimum speed limit or they're performing other unsafe maneuvers like weaving between lanes or stopping abruptly then they're clearly not a safe driver, so I agree that slow does not always equal safe.
But otherwise the collision is both morally and legally 100% the fault of the passing driver. Drivers have no moral or legal right to always operate at the speed limit if it's not safe to do so.
My semi-rural town has very narrow roads and I regularly encounter utility vehicles, school busses, farm vehicles, and cyclists on the roadway, not to mention turkeys, deer, and turtles. If a passing driver caused an accident or was pulled over and they tried to argue to the police officer that the slower vehicle "provoked them" into passing dangerously I can't imagine anyone finding that a compelling argument.
Let's say you only analyze vehicle speed data. That is clearly insufficient for training self-driving.
But it is very likely there is some correlation between speed and insurance payouts. For example, say you are able to segment your customer base including one subset with drivers who typically drive >80 miles an hour and another subset with drivers who typically drive <55 miles per hour. Are you saying that you would predict that the two populations have identical risk profiles? That speed data has no correlation with risk?
All hail AI.