What are the assumptions baked into our auto insurance policies, and how do self-driving cars challenge them? Ryan Stein from Insurance Bureau of Canada (IBC) looks at the implications that self-driving cars have on today’s auto insurance laws.

Highlights

  • In this episode of the Accenture Insurance Influencers Podcast, we speak with Ryan Stein from the Insurance Bureau of Canada (IBC).
  • Currently, humans account for 90 percent of vehicle accidents—an assumption that’s baked into auto insurance policies around the world.
  • Our current auto insurance policies aren’t equipped to deal with self-driving cars. Notably, if the auto manufacturer or technology were deemed responsible for an accident, injured parties could end up negotiating product liability insurance, which is more complex than auto insurance.
  • Auto insurance policies were challenged by the sharing economy, and insurers can learn from that experience to proactively redefine auto insurance for the arrival of self-driving cars.

Introducing the Accenture Insurance Influencers podcast

Insurance hasn’t changed much in 200 years, but everything around it has. The ground beneath insurers’ feet is shifting every day, posing challenges—and creating opportunities.

We’re excited to announce the launch of the Insurance Influencers podcast from Accenture. In season one, we address some of the big questions on insurers’ minds. How will artificial intelligence (AI) change insurance? How can insurers innovate more effectively? And how can technology enable fraud detection?

What self-driving cars mean for insurance, with Ryan Stein

Our first guest is Ryan Stein, the executive director of auto insurance policy and innovation at Insurance Bureau of Canada (IBC). First, we talked to Ryan about self-driving cars and why they don’t fit into today’s auto insurance laws. Next, Ryan discussed an IBC working paper that outlines a two-part framework for how insurers, governments and regulators can update insurance laws to accommodate self-driving cars. And finally, we looked at general principles for making sure that insurance laws are equipped to keep up with emerging technologies.

The following transcript has been edited for length and clarity.

Tell me about Insurance Bureau of Canada (IBC). What’s its role within the insurance industry in Canada?

IBC is the national trade association for Canada’s property and casualty insurance companies. We work with our members to examine the political and regulatory environment, and see if there are ways of improving it for the benefit of insurance customers across the country.

I’m looking forward to asking you about autonomous vehicles and what that means for the insurance industry. I want to start with what people mean when they talk about autonomous vehicles. I understand that there are actually five designated levels. Could you fill in our listeners who aren’t familiar with them already?

The five levels of vehicle autonomy—you can actually say that there are six, because there’s level zero—come from the Society of Automotive Engineers.

  • Level zero is no automation. The driver is in complete control of the vehicle at all times.
  • Level one has some driver assistance, like speed or cruise control.
  • Level two can take control of both the vehicle speed and lane position in some situations—for instance, on a highway.
  • Level three is limited self-driving, so the vehicle can be in full control in some situations. It can monitor the road and traffic and can also inform the driver when he or she will have to take control of the vehicle.
  • Level four is fully self-driving under certain conditions. It could be a certain area, certain weather conditions or certain roads where the vehicle can handle all the driving functions.
  • Level five is full self-driving. The vehicle can do pretty much everything without the human needing to take control.

IBC recently published a paper on what you refer to as automated vehicles. I’ve also heard the industry refer to autonomous vehicles. Are these essentially the same thing?

Yes and no. Autonomous pretty much means that the car drives itself. I like to use the word “automated” because you can talk about vehicles that still require humans to play some control in the driving operation. They have automated functions, but they might not be fully autonomous.

That brings us to the insurance industry and some of the assumptions within the insurance industry that automated vehicles may not fit into. What are some of those underlying assumptions that we’ve built into our current models of auto insurance?

The main assumption is that human error is the primary cause of collisions. The tort laws, liability laws and the liability coverage that people buy is all based on this notion that humans cause collisions. And that’s because right now, humans are responsible for over 90 percent of collisions. So it makes sense that auto insurance laws—and the coverage that comes from them—will all be based on that.

Those assumptions about auto insurance have been in place for a while and recent innovations have challenged them. So, for example, the sharing economy, ride-sharing and car-sharing. How were those a challenge to the personal auto industry?

Prior to the sharing economy, the insurance laws were written in a very specific way. Basically:

  1. A person owned a vehicle.
  2. That vehicle was predominantly used for personal or commercial purposes.
  3. The owner of that vehicle was the one who bought the coverage.

Each vehicle pretty much had one policy on it, and that policy would be personal or commercial—although you could buy optional products if you were using your vehicle for commercial purposes sometimes.

And then the sharing economy and ride-sharing services came, and it started blurring the lines between personal and commercial. People were using their vehicle for ride-sharing purposes. The ride-sharing companies wanted to be able to offer a second policy to those vehicles to cover the ride-sharing, for when the ride-sharing app is on until the ride-sharing app is off. But people that signed up for ride-sharing services didn’t really want to go out and buy a separate policy, or maybe their insurance company that sold their personal policy didn’t offer this ride-sharing policy. So for that second policy to be provided by a different entity—the ride-sharing company, not the individual vehicle owner—you needed legislative and regulatory changes.

And now, because you were going to have two policies on a vehicle, you needed rules or processes to manage claims. If a collision happened with one of those vehicles, it needed to be easy to figure out which insurance company pays. Was the app on or off? After identifying that, you could move forward with the claims process. So it was an example of insurance laws needing to be updated—to accommodate a different type of vehicle use in a different type of business model.

Right. And it strikes me that there are a lot of similarities to what we’re looking at now with automated vehicles. A lot of the conversation has been about the shift from a personal auto policy to one of product liability. Namely, if there is an accident, and it was a car that can drive itself, was it the driver or was it the manufacturer? Can you talk about some of the other implications for insurance?

Right now, humans are responsible for more than 90 percent of collisions and all the auto insurance laws and coverage is based on that. So right now, if there’s a collision, people go to their own insurance company and they get certain benefits, and if they need more and they weren’t responsible for the collision, they have an opportunity to pursue a liability claim or sue the person responsible. With motor vehicle claims, there are tens of thousands of them a year, and you figure out, OK, what the cause and was who at fault? From that, here’s how much gets paid out for the claim.

But in a world where it wasn’t the person that caused the collision—if it was the technology at fault—well, then you’re outside auto insurance litigation. Now you’re looking at product liability litigation against the vehicle manufacturer or technology provider. That’s a lot more complex and takes a lot longer than your typical motor vehicle collision liability claims.

If you have people that are injured in a collision that was caused by automated vehicle, they’ll get some coverage from their own insurer, but if they need more they’re going to have to go up against a vehicle manufacturer technology provider. It’s no longer a motor vehicle liability claim, which means that person could now be waiting a lot longer to get compensated.

And from a public policy perspective: auto insurance is heavily regulated, and at IBC we believe the laws that underpin it should make sure that people who are injured have access to fair and quick compensation. We see automated vehicles challenging the auto insurance laws that have been in place for decades, and we think there’s a need to update them. They should reflect the risks associated with automated vehicles, so you don’t have people injured having to proceed through costly, protracted product liability litigation.

That’s a great point, Ryan. Thanks for making the time to speak with me today.

It was my pleasure.

Summary

In this episode of the Accenture Insurance Influencers podcast, we talked about:

  • Six levels of driving automation, as defined by the Society of Automotive Engineers
  • The underlying assumptions baked into auto insurance policies and law, and how they were challenged by the sharing economy
  • Why today’s insurance industry isn’t prepared for automated cars, and why that should concern consumers

For more guidance on self-driving cars:

In the next episode, Ryan will share a two-part framework that IBC developed for automated vehicles and how it addresses the possibility of injured parties having to negotiate product liability insurance. And, we’ll talk about the challenges and opportunities that self-driving cars pose for insurers.

What to do next:

Contact us if you’d like to be a guest on the Insurance Influencers podcast.

Submit a Comment

Your email address will not be published. Required fields are marked *