In this hoganlovells.com interview, Dr. Sebastian Lach discusses the massive liability shift that will occur as autonomous vehicles (AV) hit the roadways and the factors that will influence how insurers calculate risk.
How will autonomous vehicles shift liability from the driver to the manufacturer?
Lach: Looking at car accidents, the question used to mainly be: Who is going to be liable — the driver or the owner of the vehicle? Product liability did play a role but a very minor one. If cars are now operated automatically there can be little doubt — if we ever come to the fact that the driver will not be acting at all — that the question or product liability will be raised. Then the question is mainly going to be: Is it a product liability issue?
The volume of car insurance cases in Germany is about US$30 billion a year. That is mainly handled by drivers’ liability or car owners’ liability, but that whole bucket of liability might move towards product liability. One of the biggest liability issues that we have in Germany and other countries today is that car insurance liability might now become a topic of product liability.
It’s less about whether the driver made a mistake, which is the issue nowadays, and is more about whether the car made a mistake. Or rather is there any other explanation — other than did the car make a mistake — if nobody else is involved? It is two automatically run cars that are going to crash — it’s two products. It’s only going to be a question about the responsibility of the product and that may go back to the original equipment manufacturer (OEM).
How will industry define safety in a world of autonomous vehicles?
Lach: Automated driving is a very innovative market with completely new products. You are never going to get perfectly safe products, but the question is: What is the state of the art? What do you have to do to say the product is reasonably safe? That’s going to be interesting to find out. Is a car reasonably safe if it is as good as a normal driver? Is a car safe if it is as good as the ideal driver? Or does an autonomous vehicle have to be better than the ideal driver because it is a machine that doesn’t have human flaws? Or is state of the art being as good as the best competitor with regard to safety? That’s going to be the discussion.
How will insurance companies calculate insurance risk as liability shifts and autonomous vehicles start to appear on roadways?
Lach: For you as a driver, the situation is going to change as follows. Imagine today you are driving on a bus. Do you need insurance? No, because you are just a passenger. That might be the same in 20 years if we talk about autonomous cars — you are just a passenger. What does that change for the liability? The whole bucket of drivers’ insurance will shift over to OEMs and insurance companies will have to decide how they insure car manufacturers.
Now we are talking about completely different amounts. Insurance is about math and insurance companies will need to think about how likely accidents are going to be. Are we going to live in a world without accidents? Are we going to live in a world with just as many accidents but not as much automated driving? There are going to be a lot of assumptions here because we are entering a completely new world. And there’s another factor that adds complexity — we are not going to have, from one day to another, only autonomous cars on the road. We are going to have a mix of traffic with people driving, others being driven, and others driving less. So it’s also going to be difficult to calculate the risks on that basis.
How is Hogan Lovells helping clients navigate this space?
Lach: We are advising clients on what sort of due diligence they have to do — testing and documentation — so that when the day comes when there is an allegation of liability, they can defend the company and also defend individuals from claims of negligence related to car accidents.
You have to monitor the market to see what others are doing so that you don’t fall behind and face an allegation that you are not state of the art. You have to carefully document that you are doing the right amount of testing. You have to make the right judgment calls when you do identify risks — that you either cure those risks with additional testing or decide that it is enough to warn about them in the product information. And you have to document why you were making these decisions and document that they were made on a sound basis based on a responsible company culture.
How can OEMs balance their desire to be first-moving with the need to conduct the necessary testing, development, and due diligence?
Lach: OEMs should try to be mindful that while there is a lot of pressure to be on the market early, before launching such a product they’ve done the necessary testing, development, and due diligence. Once the product is on the market, OEMs need to closely monitor whether it is behaving as expected. If it is not, and the OEM is seeing product risks, they need to make sure they are reacting to those risks quickly and in an appropriate manner to address them. When these products are new on the market, there is going to be a lot of scrutiny on the behavior of OEMs with regard to being responsible when it comes to dealing with product risks.
We’ve talked a lot about autonomous vehicles. What kinds of liability risks are associated with connected cars?
Lach: Connected cars will also be a problem. Now you are connecting various cars from different OEMs. If something goes wrong and there is a crash, whose fault is it? If something happened with the interface, are both the cars that collide at fault? Is the person developing the software at fault? Or are both software developers at fault? There are going to be completely new questions about responsibility that we’ve not yet faced when we talk about connected cars.
You are also going to have connected cars that are hacked. So if there is a breach into a vehicle’s software, and a hacker takes control or you have a virus through the vehicle’s connection node — is it a product flaw? Does that mean product liability because the product manufacturer didn’t prevent the hacking from happening with sufficient safeguards or is it just an attack from the outside? Or is it both? Where does that start and end?