Who’s At Fault When a Self-Driving Car Crashes?

Self-driving, or autonomous, cars are being legally tested on roads in California, Nevada, and now a section of highways in Virginia. More states will follow as various state legislatures consider bills making it legal for automakers and Google to test their vehicles on the very streets you and I drive. To date, most car and truck manufacturers who are actively working on self-driving vehicles have only tested them on public roads where legislation has permitted them to do so.

One of the questions testers, legislators, and even insurance companies are asking is who or what is at fault when a self-driving car crashes? Will it be the other driver, the autonomous car, or some combination? In addition, what testing procedures will the automaker need to follow when testing the vehicle, and whose insurance covers an accident if the self-driving technology is to blame for the accident?

Google has several self-driving cars on roads in California that have racked up 1.8 million test miles. Interestingly enough, they have been involved in 13 accidents as reported by the Wall Street Journal. According to the article, all of the accidents were not at fault accidents consisting of being side-swiped a couple of times, struck by someone who rolled through a stop sign, and rear ended by other vehicles.

In each of these cases, no one has been hurt, but the Google cars involved in the accidents have suffered some minor body damage. Since these accidents were the fault of the other driver, any property damage sustained by the Google vehicle will be covered by the other party’s car insurance as a liability claim.

Where it gets murky, is what if the accident had been the fault of the Google vehicle, or any other autonomous car or truck. Any car or truck manufacturer’s commercial general liability or product liability policy will most likely pay for bodily injury and property damage claims if one of their vehicles is the cause of the accident.

Insurance companies are already being forced to address these issues. Many cars are available today with a variety of driver assist technology including lane departure warning, brake assist, and even parking assist. In January of 2014, a Florida driver engaged the parking assist function of his BMW at a local Target. The car accelerated, climbed a small tree, and mounted the car parked across from it. BMW asserted the parking assist does not control the car’s acceleration, which is the responsibility of the driver. The Sheriff’s department’s initial thought was the driver may have mistakenly hit the accelerator instead of the brake.

In a case like this, if the technology malfunctioned and was determined to have caused the wreck, BMW’s product liability insurance would ultimately pay for the damage to both cars. If it was determined the driver was at fault and pressed on the accelerator instead of the brake pedal, then the car owner’s insurance will be tapped to pay for damages to the vehicles. Insurers will have an interesting time reviewing and analyzing a vehicle’s computer data to determine if the error was caused by a computer malfunction or the driver. It makes me wonder how different are the computer systems in each car and truck manufacturer, or is there some level of continuity similar to a black box found on all commercial airliners? A potentially bigger issue, is how will these systems ward off viruses or prevent being hacked? No one wants to hit control – alt – delete while driving!

The promise of autonomous cars is very exciting for a large metropolitan area like Dallas / Fort Worth, Houston, etc. I believe, that at least until these vehicles become the standard, car insurance will still be required until the technology is proven. What do you think? Share your thoughts, questions, and experiences with me on our Google +, Facebook, and LinkedIn pages. I’d love to hear from you!

Evie Wise
Evie Wise


Evie Wise
Evie Wise

Share this post with your friends