Autonomous cars: who's at fault when they go wrong?


Everyone has seen the movies from the 1980’s and 90’s which predicted we would all be travelling around in vehicles that hovered or flew in the future. The one aspect that remained in all of these is that they were still piloted by a being behind the wheel. None seemed to contemplate that the vehicles would be computer driven; although, unfortunately, still on four wheels rather than hovering!

There are many unanswered questions as we seem to be heading towards an autonomous car world, one of which is whether we would really be safer with a computer driving, thus reducing the amount of accidents on the road. The second question is, when an accident does happen, who is at fault?

From anyone’s perspective this is an interesting topic but as a Personal Injury lawyer who has dealt with hundreds of accidents involving vehicles, there is an added dimension to my views and thoughts.

No one wants to hear of anyone being injured in a road accident, however when it does happen there is currently a definite Defendant in the form of the driver of the vehicle at fault. If you take away the human element, then who would be to blame for an accident if it involved autonomous vehicles?

Road tests for autonomous vehicles have shown that the system is not without fault. Early in 2016 an autonomous vehicle pulled into the path of a bus being driven by a person and caused a collision. It had not been calibrated to be able to differentiate as to whether the bus was slowing down to stop or accelerating to get going. The accident occurred at around 15 mph but, given it was a bus, there was significant damage to the autonomous vehicle and, had there been an occupier of the vehicle, I am sure they would have suffered some form of injury. Therefore, can a car, without a human pilot adapt to its surroundings and react in time to prevent a collision, especially where the other vehicle is being piloted by a human? Further, autonomous cars, like any computer programme, are only as good as the programmer who writes the software; more specifically, the various different scenarios which they imagine when writing the same. It may well take some time for each and every possible scenario to come to light and for software to cover all such eventualities. At the moment I would suggest it would be difficult to be 100% confident that a computer could adapt in time.

So who could be at fault when an accident does occur?

With some vehicles being spoken about as having no over-riding controls, the vehicle’s occupant cannot be seen as being at fault as they were not driving the vehicle. The logical argument would push liability back to the manufacturer, as it was their computer system that was at fault. Does the issue suddenly stop being a road traffic accident and become a product liability case for the provision of a faulty product? These carry a different set of rules from a Road Traffic Accident.It comes down to not only time limits but also the costs payable by a Defendant as well as the issues to be proved by a potential Claimant (as there is still strict liability in product liability cases).

There is also the question of how well the computer would adapt to cyclists, motorcyclists and pedestrians, all of which carry far more serious injuries in a like for like collision. This is before you start to look into whether the computer can be hacked and controlled by an outside influence - but that is another topic altogether!

There seem to be far more questions than answers at this point in time. However, there are whispers that autonomous cars could hit the roads within the next 10-20 years despite these questions. Is this enough time to ensure full road safety? No one seems to be providing answers. Would you be happy putting your life in the hands of a computer? Are we really ready for an autonomous car future? I know what my answer is…