States haven’t said whether a manufacturer or passenger is responsible in a driverless-car crash, so insurance markets can’t price the risk. “We don’t yet know how judges and juries will react,” says Bryant Walker Smith of Stanford Law School. That’s one barrier to widespread adoption.
Enact laws that make the manufacturer liable. That way, insurance could be bundled with the car when it’s sold, and rates would reflect the manufacturer’s driving record, not the passengers’, since self-driving cars would be “closed” robots—humans can’t override them. Although that may sound alarming, it’s the same thinking behind public transportation: you surrender control when you get on a bus, so you’re not liable if something goes wrong. (And the chances of something going wrong are much, much smaller than if you’re driving yourself.) Barring that, “insurance will simply cost too much,” says former acting U.S. solicitor general Neal Katyal, who has advised Google on the same issue.
Robot cars could reduce the more than 30,000 annual deaths from crashes, nearly a third of which are alcohol-related. The elderly and disabled would benefit, and pollution would drop.
MORE: Regulate the Robots