Insurance could be the biggest roadblock for autonomous cars

Even if self-driving cars take off, robots and humans sharing the roads remains a complicated problem for insurers

Google’s prototype autonomous car.

Google’s prototype autonomous car. (Google)

The robot cars are coming for us, right? Well, maybe. There’s the small problem of auto insurance to deal with first.

Companies such as Google are promising commercially available autonomous vehicles by 2020, but insurance realities could slow that down considerably. After all, what happens when a robot car slams into a human-driven one—who’s responsible?

“Until everyone on the road is driving a completely autonomous self-driving car, I’m not sure that anyone could be allowed to drive a self-driving car,” says Jason Katz, a personal injury lawyer at Toronto firm Singer Kwinter.

“Even if the vehicle is doing the majority of the driving, if the human has to step in to prevent an accident and stay focused on the road, then it’ll be the driver who is responsible.”

It’s a dichotomy that hasn’t yet received much consideration. Autonomous vehicles hold much promise in making driving safer, and existing technological improvements and additions already seem to be having a positive effect on car fatalities. The true benefits of robot cars, however, are only likely to accrue when they are the norm rather than the exception.

Consider the 2012 study out of Columbia University that suggested highway capacity could triple if robot cars become widespread. Autonomous cars would be able to safely drive in closer proximity to each other, at faster speeds. All it would take to ruin that fast-moving swarm is one foolish, poorly driving human.

The same would likely hold true in insurance cases, Katz says. As long as there are some human drivers on the road, the issue of liability will continue to rage.

“Until every car on the road is completely self-driving the human factor would still have to be a part of it and the buck stops with the human behind the wheel.”

One possibility, he suggests, is a switch-over to a system of absolute liability, where insurance is sold bundled with cars. There would no longer be a question of fault—the manufacturer of the robot car would simply automatically be liable for any damages in an accident.

Self-driving car makers are likely to oppose such a suggestion, but it wouldn’t necessarily have to be as bad as it sounds for them. Another possible model floated a few years ago would be similar to vaccines. In the United States, vaccine makers pay into a trust fund that compensates people who fall ill to them.

Robot car makers could do the same, at least until wide-scale conversion to the vehicles was completed.

Another possibility could see governments fast track initiatives to make robot cars mandatory, although such wide-scale rollouts would still likely take years.

So far, there aren’t any clear and simple solutions. Katz sums it up with an understatement: “There are some major hurdles here.”